Compare commits

...

687 Commits

Author SHA1 Message Date
Tibor Vass 14680bf724 Bump version to v0.12.0
Docker-DCO-1.1-Signed-off-by: Tibor Vass <teabee89@gmail.com> (github: tiborvass)
2014-06-05 17:59:19 -07:00
Tibor Vass c06f32a7b9 Merge pull request #6228 from vieux/fix_restart_links
Fix restart links
2014-06-05 17:54:27 -07:00
Victor Vieux 30ba7546cb add wait4 after kill
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-06 00:32:14 +00:00
Victor Vieux 34bd2d6229 start containers after all of them are registered
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-06 00:31:58 +00:00
James Turnbull 98d44ac491 Merge pull request #6226 from fredlf/copyedits
More minor copy edits, typo fix.
2014-06-05 17:07:18 -07:00
Tianon Gravi c6dc8275b2 Merge pull request #6227 from crosbymichael/fix-test-compile
Don't compile nsinit during tests (it has package main)
2014-06-05 18:06:31 -06:00
Michael Crosby 3986bc659b Don't compile nsinit
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-05 17:03:32 -07:00
Fred Lifton 762c528501 More minor copy edits, typo fix.
Docker-DCO-1.1-Signed-off-by: Fred Lifton <fred.lifton@docker.com> (github: fredfl)
2014-06-05 16:55:32 -07:00
Michael Crosby acb9f156fd Merge pull request #6207 from crosbymichael/nsinit-cli
Make nsinit nicer to work with libcontainer
2014-06-05 16:51:31 -07:00
Michael Crosby c239c8ad9d Merge pull request #6221 from crosbymichael/no-unmount
Move remount as private to the graph drivers
2014-06-05 16:51:10 -07:00
Michael Crosby 3609b051b8 Move remount as private to the graph drivers
If this is at the root directory for the daemon you could unmount
somones filesystem when you stop docker and this is actually only needed
for the palces that the graph drivers mount the container's root
    filesystems.
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-05 16:02:25 -07:00
Tibor Vass 9329c0d2e0 Merge pull request #6224 from tiborvass/storage-options
Add --storage-opt daemon option and some devicemapper option (with fixes)
2014-06-05 16:00:03 -07:00
unclejack 2f72fdf6ea Merge pull request #6145 from shykes/pr_out_pkg_testutils_utility_functions_to_facilitate_writing_go_tests
pkg/testutils: utility functions to facilitate writing Go tests
2014-06-06 01:26:53 +03:00
James Turnbull 19ef3ff112 Merge pull request #6182 from mheon/fix_manpage_issue
Clarify manpages on docker start
2014-06-05 15:15:50 -07:00
Tibor Vass b1ac791d84 Ensures files get closed properly. Closes #6213
Docker-DCO-1.1-Signed-off-by: Tibor Vass <teabee89@gmail.com> (github: tiborvass)
2014-06-05 15:13:43 -07:00
Sven Dowideit a2c553bd6b Merge pull request #6219 from cpuguy83/fix_ssh_example_docs_root_login
Fixes broken ssh server example
2014-06-05 15:02:43 -07:00
Sven Dowideit 64ce33268c Merge pull request #6222 from SvenDowideit/pr_out_reflow_to_80_columns
Initial edits to home page
2014-06-05 14:58:29 -07:00
SvenDowideit abcb724ed9 reflow to 80 columns
Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-06-06 07:58:05 +10:00
Fred Lifton 09b19752e0 Initial copy edits to home page.
Docker-DCO-1.1-Signed-off-by: Fred Lifton <fred.lifton@docker.com> (github: fredlf)

Docker-DCO-1.1-Signed-off-by: Fred Lifton <fred.lifton@docker.com> (github: SvenDowideit)
2014-06-06 07:58:04 +10:00
Sven Dowideit 6be75947ab Merge pull request #6223 from jamtur01/rl
Revert "Initial links for Docker Hub rename"
2014-06-05 14:56:12 -07:00
James Turnbull fad14eda3a Revert "Initial links for Docker Hub rename"
This reverts commit 2819677c21.

Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-06-05 14:52:06 -07:00
James Turnbull 6fa3acc96c Merge pull request #6156 from jamtur01/links
Initial link update for Docs
2014-06-05 14:46:33 -07:00
Victor Vieux cedab4d06a Merge pull request #6220 from vieux/fix_redirect
Only forward auth to trusted locations
2014-06-05 14:34:37 -07:00
Michael Crosby f362153157 Update for nsenter
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-05 14:31:16 -07:00
Michael Crosby 1a3d43c23e Update nsinit to be nicer to work with and test
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-05 14:13:02 -07:00
Victor Vieux a124531863 only forward auth to trusted locations
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-05 19:47:33 +00:00
Matthew Heon d08c965e2c Clarify effect of docker start on started container in CLI docs
This behavior changed from v0.9 to v0.10, so document it to prevent
any confusion.

Docker-DCO-1.1-Signed-off-by: Matthew Heon <mheon@redhat.com> (github: mheon)
2014-06-05 15:08:54 -04:00
cpuguy83 04620e0154 Fixes broken ssh server example
Docker-DCO-1.1-Signed-off-by: cpuguy83 <cpuguy83@gmail.com> (github: cpuguy83)
2014-06-05 14:10:52 -04:00
Sven Dowideit 734d108ed2 Merge pull request #6215 from cpuguy83/fix_volumes_from_api_doc
Fixes incorrect API spec for items moved to HostConfig
2014-06-05 10:33:16 -07:00
cpuguy83 8c646ebf16 Fixes incorrect API spec for items moved to HostConfig
DNS and VolumesFrom were moved to HostConfig and as such are part of the
container start and not create.
For some reason 0.10 docs are correct (except for a missing quote in the
JSON") but 0.11 and latest are not.

Docker-DCO-1.1-Signed-off-by: cpuguy83 <cpuguy83@gmail.com> (github: cpuguy83)
2014-06-05 09:50:27 -04:00
Alexander Larsson 0434a2ce64 devmapper: Add blkdiscard option and disable it on raw devices
The blkdiscard hack we do on container/image delete is pretty slow, but
required to restore space to the "host" root filesystem. However, it
is pretty useless on raw devices, and you may not need it in development
either.

In a simple test of the devicemapper backend on loopback the time to
delete 20 container went from 11 seconds to 0.4 seconds with
--storage-opt blkdiscard=false.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-06-05 11:45:59 +02:00
Alexander Larsson a226168a8b devmapper: Add options for specifying block devices
This adds dm.datadev and dm.metadatadev options that you can use with
--storage-opt to set to specific devices to use for the thin
provisioning pool.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-06-05 11:45:55 +02:00
Alexander Larsson 807bc2cd04 devmapper: Allow specifying filesystem for thin devices
This adds the following --storage-opts for the daemon:
dm.fs: The filesystem to use for the base image
dm.mkfsarg: Add an argument to the mkfs command for the base image
dm.mountopt: Add a mount option for devicemapper mount

Currently supported filesystems are xfs and ext4.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-06-05 11:25:06 +02:00
Alexander Larsson 10083f4140 devicemapper: Probe what filesystem to use when mounting
Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-06-05 11:03:19 +02:00
Alexander Larsson 7f5ba068f4 devmapper: Add --storage-opt options for basic devicemapper settings
This allows setting these settings to be passed:
dm.basesize
dm.loopdatasize
dm.loopmetadatasize

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-06-05 10:50:07 +02:00
Alexander Larsson 822ea97ffc Add --storage-opt graph driver option and pass through to driver
This lets you add storage specific options for the daemon.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-06-05 10:42:27 +02:00
Alexander Larsson 948e54ac45 devmapper: Fail init with ErrNotSupported if simple devmapper call fails
If we can't even get the current device mapper driver version, then
we cleanly fail the devmapper driver as not supported and fall back
on the next one.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-06-05 10:42:27 +02:00
Alexander Larsson 13f07b636f pkg/units: Add FromHumanSize
This does the "reverse" of HumanSize, i.e. maps a string to an int64
using SI prefixes for the extension.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-06-05 10:02:37 +02:00
Victor Vieux 7e96f8de1c Merge pull request #6026 from vmarmol/nsenter
Adding initial version of C-based nsenter for allowing execin in libcontainer.
2014-06-04 19:06:27 -07:00
Victor Vieux fe706af3c0 Merge pull request #6210 from tiborvass/fix_doc_remove_insert_remote_api
Move doc section in Remote API about removing insert from v1.11 to v1.12
2014-06-04 18:31:55 -07:00
Tibor Vass 0d1906feea Move doc section in Remote API about removing insert from v1.11 to v1.12
Docker-DCO-1.1-Signed-off-by: Tibor Vass <teabee89@gmail.com> (github: tiborvass)
2014-06-04 18:30:10 -07:00
Tibor Vass 894814f708 Merge pull request #6208 from crosbymichael/remove-insert
Totally remove insert feature
2014-06-04 18:17:23 -07:00
Victor Marmol a0137e104c Merge pull request #1 from crosbymichael/nsenter-path
Move environment loading to nsenter
2014-06-04 18:14:12 -07:00
Michael Crosby 7eb508633d Exclude the user namespace for setns
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-04 18:03:17 -07:00
Michael Crosby 8497d1274b Move env load to nsenter
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-04 17:54:00 -07:00
Victor Marmol 0a725ea282 Adding initial version of C-based nsenter for allowing execin in
libcontainer.

Docker-DCO-1.1-Signed-off-by: Victor Marmol <vmarmol@google.com> (github: vmarmol)
2014-06-05 00:44:13 +00:00
Michael Crosby 7a145b022a Totally remove insert feature
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-04 17:35:15 -07:00
Michael Crosby a8ef66d84a Merge pull request #6205 from crosbymichael/rename-namespaces
Rename nsinit package to namespaces in libcontainer
2014-06-04 17:27:45 -07:00
James Turnbull 2819677c21 Initial links for Docker Hub rename
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-06-04 16:48:49 -07:00
Victor Vieux aa05170a82 Merge pull request #6204 from crosbymichael/no-structs
Remove the API container struct and use env
2014-06-04 16:40:20 -07:00
Michael Crosby 8aff01c0b4 Rename nsinit package to namespaces in libcontainer
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-04 15:47:57 -07:00
Michael Crosby 91757eaeb3 Remove the API container struct and use env
This change touches docker logs, attach, and start -a along with get
exit code in the client.
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-04 15:37:49 -07:00
unclejack 048833c246 Merge pull request #6178 from calfonso/freezer-pr-integration
Freezer integration to support container pause/unpause
2014-06-05 01:29:47 +03:00
Michael Crosby fd21da4b0b Merge pull request #6198 from vishh/stats2
Add more stats to libcontainer.
2014-06-04 15:28:19 -07:00
unclejack 597c06ca34 Merge pull request #6195 from kieslee/master
pass address of metaHeaders to GetenvJson in ImagePush and ImagePull
2014-06-05 00:23:08 +03:00
Vishnu Kannan 8d63b610c3 Add stats for memory allocation failure count and instantaneous cpu usage in the usermode and kernelmode.
Docker-DCO-1.1-Signed-off-by: Vishnu Kannan <vishnuk@google.com> (github: vishh)
2014-06-04 21:20:21 +00:00
Chris Alfonso 26246ebd53 Integrating systemd freeze functionality.
This pulls together #6061 and #6125

Docker-DCO-1.1-Signed-off-by: Chris Alfonso <calfonso@redhat.com> (github: calfonso)
2014-06-04 13:33:44 -06:00
Ian Main b054569cde Add ability to pause/unpause containers via cgroups freeze
This patch adds pause/unpause to the command line, api, and drivers
for use on containers.  This is implemented using the cgroups/freeze
utility in libcontainer and lxc freeze/unfreeze.

Co-Authored-By: Eric Windisch <ewindisch@docker.com>
Co-Authored-By: Chris Alfonso <calfonso@redhat.com>
Docker-DCO-1.1-Signed-off-by: Ian Main <imain@redhat.com> (github: imain)
2014-06-04 13:33:44 -06:00
Michael Crosby 1f5b323208 Merge pull request #6187 from vieux/wait_freeze
Implement wait on freeze
2014-06-04 12:00:20 -07:00
kies eb3e2f4f47 pass address of metaHeaders to GetenvJson in ImagePush and ImagePull
Docker-DCO-1.1-Signed-off-by: kies <lleelm@gmail.com> (github: kieslee)
2014-06-04 17:54:06 +08:00
Sven Dowideit 9f14ff83b2 Merge pull request #6191 from snowch/patch-1
Small clarification
2014-06-04 01:56:01 -07:00
Chris Snow 2e14f0f70a Update builder.md 2014-06-04 08:08:44 +01:00
Victor Vieux 3704996089 Merge pull request #6082 from FLGMwt/2342-change-wording-uploading-context
Change 'Uploading context' wording
2014-06-03 23:58:20 -07:00
Victor Vieux dc21dd2634 implement wait on freeze
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-04 02:21:10 +00:00
Tibor Vass 2a92c53a06 Merge pull request #6186 from unclejack/fix_typo_aufsgo
fix diffs->diff typo in aufs.go
2014-06-03 16:59:43 -07:00
unclejack 7a60b9063c fix diffs->diff typo in aufs.go
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-06-04 02:46:55 +03:00
Victor Vieux 2c27e6e9ab Merge pull request #6183 from crosbymichael/fix-freezer-fs
Don't rejoin the freezer cgroup each time
2014-06-03 16:16:58 -07:00
Victor Vieux 37d69aae5e Merge pull request #6180 from unclejack/apparmor
apparmor: write & load the profile on every start
2014-06-03 15:56:28 -07:00
Michael Crosby 534521301c Don't rejoin the cgroup each time
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-03 15:24:30 -07:00
unclejack a2ec8fa1db Merge pull request #6181 from SvenDowideit/change-out-port-number
IANA allocated Docker port: 2375
2014-06-04 01:07:48 +03:00
Sven Dowideit 201d20fe34 Merge pull request #6164 from jamtur01/automated
Changed all references from Trusted Builds to Automated Builds
2014-06-03 15:06:09 -07:00
unclejack 1ef3ca83d8 apparmor: write & load the profile on every start
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-06-04 00:56:35 +03:00
Matthew Heon d96eb47dee Added note in manpages to clarify effect of docker start on started container
When run on a container that is already running, docker always succeeds.
This was an intentional change in v0.10; it previously caused an error.
This commit clarifies the man pages to indicate this is intended behavior.

Docker-DCO-1.1-Signed-off-by: Matthew Heon <mheon@redhat.com> (github: mheon)
2014-06-03 17:10:59 -04:00
SvenDowideit 5febba93ba IANA allocated Docker port: 2375
2375/2376 are assigned:
http://www.iana.org/assignments/service-names-port-numbers/service-names-port-numbers.xhtml?search=docker

For #1440

Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-06-04 06:54:19 +10:00
James Turnbull 9fc8028c98 Changed all references from Trusted Builds to Automated Builds
* Updated docker images output
* Deprecated docker images -t/--trusted option and replace with --automated
* Changed *trusted variables to *automated
* Changed added support for is_automated alongside deprecated is_trusted
* Updated man pages, docs and completion
* Updated API documentation

Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-06-03 12:55:30 -07:00
Victor Vieux 2f514fb891 Merge pull request #6125 from crosbymichael/systemd-freezer
Implement systemd support for freezer
2014-06-03 11:39:21 -07:00
Sven Dowideit 7f55d92837 Merge pull request #6173 from zdover23/03_Jun_2014
grammar; nothing big (added a "the")
2014-06-03 10:48:20 -07:00
Michael Crosby 613f74c1fb Implement systemd support for freezer
These PR does a few things.  It ensures that the freezer cgroup is
joined in the systemd driver.  It also provides a public api for setting
the freezer state via the cgroups package.
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-03 10:46:53 -07:00
Michael Crosby ae93939414 Merge pull request #6175 from alexlarsson/systemd-use-raw-devices
libcontainer/cgroup: Use raw access to set up and join the devices cgroup
2014-06-03 10:38:31 -07:00
unclejack 5aeb77296a Merge pull request #4430 from vbatts/vbatts-images_orphan
Adding -f untagged=true flag to `docker images`
2014-06-03 20:33:56 +03:00
Vincent Batts 02255ddaa4 filters: don't encode empty set. API docs
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-03 12:19:26 -04:00
Alexander Larsson 6b8f0e394b libcontainer/cgroup: Use raw access to set up and join the devices cgroup
The systemd support for the devices cgroup lacks two required features:
 * Support for wildcards to allow mknod on any device
 * Support for wildcards to allow /dev/pts support

The second is available in more recent systemd as "char-pts", but not in e.g. v208 which is in wide use.

Additionally, the current approach of letting systemd set up the devices cgroup and then adding
some devices to it doesn't work, because some times systemd (at least v208) re-initializes
the devices cgroup, overwriting our custom devices. See https://github.com/dotcloud/docker/issues/6009
for the details.

When wildcarded mknod support is available in systemd we should implement a pure systemd version,
but we need to keep the old one around for backwards compat.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-06-03 11:34:14 +02:00
Victor Vieux dccf336204 Merge pull request #6172 from tianon/remove-duplicated-magic-values
Move duplicated FS "magic" values to the graphdriver package so they can be shared
2014-06-02 23:25:46 -07:00
Zac Dover 3a77a73a85 grammar; nothing big
Docker-DCO-1.1-Signed-off-by: Zac Dover <zdover@redhat.com> (github: zdover23)

Added another "The"

Docker-DCO-1.1-Signed-off-by: Zac Dover <zdover@redhat.com> (github: zdover23)
2014-06-03 15:40:22 +10:00
James Turnbull 19e4fddcce Merge pull request #6122 from mheon/fix_manpage_issue
Added documentation for special case in run command
2014-06-02 21:19:43 -07:00
Tianon Gravi 68476e277f Move duplicated FS "magic" values to the graphdriver package so they can be shared instead of duplicated
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-06-02 19:56:47 -06:00
Victor Marmol 5bf4068d60 Merge pull request #6171 from crosbymichael/add-chroot
Add SYS_CHROOT cap to unprivileged containers
2014-06-02 18:33:34 -07:00
Michael Crosby 41f7cef2bd Add SYS_CHROOT cap to unprivileged containers
Fixes #6103
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-02 18:23:47 -07:00
unclejack 3fb0870878 Merge pull request #6167 from crosbymichael/volume-ownership
Retain volume ownership and permissions
2014-06-03 04:23:30 +03:00
Victor Vieux fd342cb777 Merge pull request #6170 from vieux/redirect_login
Add redirect and env proxy support to docker login
2014-06-02 18:19:28 -07:00
Michael Crosby c024c9bd1e Add test for volume ownership and perms
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-02 18:10:52 -07:00
Michael Crosby f41ced96af Ensure that ownership and perms are copied to volume
This only works if the file or dir is already created in
the image before setting it to be a volume.  There is no way around this
because we don't have the data avaliable to set the volume at the
beginning of the dockerfile
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-02 18:06:58 -07:00
Michael Crosby f65fadbda0 Merge pull request #6143 from LK4D4/move_some_more_tests_to_cli
Move some more tests to integration cli
2014-06-02 18:06:11 -07:00
Sven Dowideit dd93845ad0 Merge pull request #6056 from ipbabble/master
Made fixes to man pages to remove some ambiguities and typos
2014-06-02 18:01:22 -07:00
unclejack a8ade72b6a Merge pull request #6168 from vieux/add_proto_validation
add proto validation at parse
2014-06-03 03:55:41 +03:00
Victor Marmol 70b778751e Merge pull request #6153 from vishh/stats1
Add per cpu usage to libcontainer  stats
2014-06-02 17:53:17 -07:00
Victor Vieux 4e7254840b Add redirect and env proxy support to docker login
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-03 00:46:06 +00:00
Victor Vieux 98d7b720c2 Merge pull request #6169 from vieux/add_volumes_inspect_back
add volumes back to inspect
2014-06-02 17:07:00 -07:00
Victor Vieux 996133b9ba add volumes back to inspect
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-03 00:05:54 +00:00
Victor Vieux 0633b12b28 add proto validation at parse
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-02 23:03:10 +00:00
Victor Vieux b292928cd5 Merge pull request #5848 from unclejack/resumable_pulls
resume pulling the layer on disconnect
2014-06-02 14:07:01 -07:00
Victor Vieux dcba54aa99 Merge pull request #6166 from vieux/debug_skipping
switch skipping from error to debug
2014-06-02 14:01:49 -07:00
Victor Vieux af6ab357e8 switch skipping from error to debug
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-02 20:47:07 +00:00
unclejack aa61cc759b Merge pull request #6165 from vieux/update_tests
Update tests regarding busybox & inspect
2014-06-02 23:39:12 +03:00
Vincent Batts 3a4e3ca327 filters: don't encode an empty set. update comments
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:52 -04:00
Vincent Batts 89a15fa235 filters: use json marshal, instead of beam/data
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:52 -04:00
Vincent Batts 3ecfaa8f2d filters: change untagged to dangling
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:52 -04:00
Vincent Batts f322168563 server: very debug.
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:52 -04:00
Vincent Batts f1cc7ce5d7 filter flag: split out for separate --filter flags
adding tests and allowing for easy passing of filters.Args from client
to server.

Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:51 -04:00
Vincent Batts 3391aac2b2 images: fix markdown documentation
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:51 -04:00
Vincent Batts cb7857de5d images filter: remove the redundant inverted filter
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:51 -04:00
Vincent Batts babd572015 filters: cleanup & fmt
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:51 -04:00
Vincent Batts 55d95185ed filters, for images: docs
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:51 -04:00
Vincent Batts caf9b19b0c filters: remove out filter proc prototype
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:50 -04:00
Vincent Batts 5f3812ec97 filters, for images: start with untagged/tagged boolean
This is a new feature and flag. (replaces the suggestion of a flag for
--untagged images).
The concept is to have a syntax to filter. This begins with this
filtering for the 'images' subcommand, and at that only filtering for
whether images are untagged.
  example like: docker rmi $(docker images -q --filter 'untagged=true')

Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-06-02 16:33:50 -04:00
Victor Vieux 730d9ba174 fix busybox image detection
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-02 19:57:20 +00:00
Victor Vieux b8932abcd3 pull only busybox:latest
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-06-02 19:54:17 +00:00
LK4D4 ae128437ce Move build cache tests to integration-cli
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-06-02 23:40:14 +04:00
LK4D4 bf4d907092 More verbose build tests
I've decided that custom asserts only hide the meaning of tests
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-06-02 23:40:13 +04:00
LK4D4 2e85568816 Aux functions for build testing
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-06-02 23:40:13 +04:00
Sven Dowideit afb2d5de3d Merge pull request #6159 from SvenDowideit/s3-json-fixup
aws json is not javascript style json. trailing comma breaks it :(
2014-06-02 12:28:57 -07:00
Victor Vieux 55230b3332 Merge pull request #6138 from NERSC/6092-setuidrace
Fix for setuid race condition in LXC driver
2014-06-02 12:16:00 -07:00
Victor Vieux c136384b20 Merge pull request #6160 from crosbymichael/update-test-for-nil-error
Update cpu stat test for no error
2014-06-02 12:12:49 -07:00
Victor Vieux 24ea088124 Merge pull request #6147 from cyphar/6142-mflag-strip-quotes-long-equal-form
mflag: strip quotes in certain forms from flag values
2014-06-02 12:03:58 -07:00
Michael Crosby 3e13aaec00 Merge pull request #6130 from vieux/standardize_api_keys
Standardize api keys to CamelCase
2014-06-02 12:03:11 -07:00
Victor Vieux 43b7af1dd1 Merge pull request #6150 from LK4D4/remove_used_from_streamformatter
Remove unused and racy "used" param from streamformatter
2014-06-02 11:56:05 -07:00
Michael Crosby 30fdac2cb1 Update cpu stat test for no error
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-06-02 11:54:23 -07:00
Michael Crosby f1a217be0f Merge pull request #6116 from vishh/nsinit_stats
Adding "stats" and "spec" option to nsinit CLI
2014-06-02 11:44:37 -07:00
unclejack 2a1b7f222a resume pulling the layer on disconnect
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-06-02 21:41:22 +03:00
SvenDowideit cac24bbfaf aws json is not javascript style json. trailing comma breaks it :(
Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-06-03 04:30:32 +10:00
James Turnbull d3ff00948f Merge pull request #6157 from SvenDowideit/pr_out_update_networking_md
Update networking.md
2014-06-02 13:46:26 -04:00
Alexandr Morozov 1ae37cef91 Remove unused and racy "used" param from streamformatter
Also tests written
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-06-02 21:43:07 +04:00
Andre Dublin 4ac216d8d3 Update networking.md
grammar

Docker-DCO-1.1-Signed-off-by: Andre Dublin <81dublin@gmail.com> (github: andredublin)

rebased by

Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-06-03 03:41:53 +10:00
Sven Dowideit 2404ce3d26 Merge pull request #6154 from SvenDowideit/pr_out_adding_user_guide
Adding User Guide
2014-06-02 10:39:15 -07:00
Solomon Hykes 0a06e9bd91 Fix format in maintainers files
Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-06-02 07:05:06 +00:00
Solomon Hykes 1a93d3b054 Add Cristian as maintainer to pkg/testutils
Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-06-02 07:01:17 +00:00
Vishnu Kannan 3723d6341e Adding percpu usage to cgroup stats reported by libcontainer.
Docker-DCO-1.1-Signed-off-by: Vishnu Kannan <vishnuk@google.com> (github: vishh)
2014-06-02 06:56:15 +00:00
Vishnu Kannan f7b82be0dd Adding "stats" and "spec" option to nsinit binary which will print the stats and spec respectively.
Docker-DCO-1.1-Signed-off-by: Vishnu Kannan <vishnuk@google.com> (github: vishh)
2014-06-02 06:37:22 +00:00
cyphar e4497feaba pkg: mflag: flag: added tests for quote-stripped flags
This patch adds some tests to ensure that quoted flags are properly
handled by the mflag package.

Docker-DCO-1.1-Signed-off-by: Aleksa Sarai <cyphar@cyphar.com> (github: cyphar)
2014-06-02 15:25:38 +10:00
cyphar 0e9c40eb82 pkg: mflag: flag: make mflag strip quotes in -flag="var" forms
This patch improves the mflag package to ensure that things arguments
to mflag such as `-flag="var"` or `-flag='var'` have the quotes
stripped from the value (to mirror the getopt functionality for similar
flags).

Docker-DCO-1.1-Signed-off-by: Aleksa Sarai <cyphar@cyphar.com> (github: cyphar)
2014-06-02 15:25:33 +10:00
Solomon Hykes ca231b3de5 pkg/testutils: utility functions to facilitate writing Go tests
Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-06-01 23:23:31 +00:00
William Henry 2858180a9a Changed the term rename to alias etc.
Docker-DCO-1.1-Signed-off-by: William Henry <whenry@redhat.com> (github: ipbabble)

 Changes to be committed:
	modified:   contrib/man/md/docker-tag.1.md
2014-06-01 17:16:14 -06:00
James Turnbull a7b2c4804b Adding User Guide
* Added User Guide section outlines.
* Added User Guide to menu.
* Moved HTTPS example to articles.
* Replaced Hello World example with User Guide.
* Moved use cases out of examples.
* Updated Introduction to add User Guide.
* Redirected migrated /use and /articles links.
* Added Docker.io section
* Added Dockerized section
* Added Using Docker section
* Added Docker Images section
* Added Docker Links section
* Added Docker Volumes section

Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-06-01 17:34:07 -04:00
Sven Dowideit 4edcbfdeb7 Merge pull request #6140 from zanedeg/patch-1
Updated link to new docker-node-hello repository
2014-06-02 07:32:43 +10:00
Sven Dowideit 758e23df4d Merge pull request #5924 from SvenDowideit/docs-build-instructions
Add documentation build steps for the next release manager to follow.
2014-06-01 08:31:51 +10:00
SvenDowideit 953abf6ae7 Add documentation build steps for the next release manager to follow.
Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-06-01 08:31:33 +10:00
unclejack 2024a0e517 Merge pull request #6089 from unclejack/docker_build_copy
add support for COPY to docker build
2014-06-01 01:26:31 +03:00
William Henry e471a87f19 Fixed some typos and other issues from ostezer comments.
Docker-DCO-1.1-Signed-off-by: William Henry <whenry@redhat.com> (github: ipbabble)

 Changes to be committed:
	modified:   contrib/man/md/docker-build.1.md
	modified:   contrib/man/md/docker-run.1.md
	modified:   contrib/man/md/docker-tag.1.md
2014-05-31 15:44:17 -06:00
O.S. Tezer b51802a47c Merge pull request #6073 from SvenDowideit/b2d-port-redirection-docs
sync the initial port redirection docs from b2d
2014-05-31 13:58:28 -07:00
Zane DeGraffenried 0658868c7f Updated link to new docker-node-hello repository 2014-05-31 14:55:59 -06:00
Shane Canon f9705477d0 Fix for setuid race condition in LXC driver
This is a fix for a race condition in the LXC driver.  This is described
more in issue #6092.

Closes #6092

Docker-DCO-1.1-Signed-off-by: Shane Canon <scanon@lbl.gov> (github: scanon)
2014-05-31 10:42:48 -07:00
unclejack 05cf3498a6 Merge pull request #6128 from crosbymichael/empty-sqlite
Init database if empty file
2014-05-31 06:56:00 +03:00
unclejack af31fe95cd Merge pull request #6129 from crosbymichael/copy-all-in-priv
Ensure all dev nodes are copied for privileged
2014-05-31 06:40:20 +03:00
Michael Crosby 69989b7c06 Ensure all dev nodes are copied for privileged
This also makes sure that devices are pointers to avoid copies
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-30 18:39:11 -07:00
Victor Vieux 1dd02ff4a0 update docs
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-31 01:28:08 +00:00
Victor Vieux 68fb7f4b74 Standardize API keys: CamelCase
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-31 01:22:07 +00:00
Michael Crosby b0ea389c69 Init database if empty file
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-30 17:44:00 -07:00
unclejack 0ef637722f Merge pull request #6097 from timthelion/consistentdevices
Refactor device handling code
2014-05-31 03:34:52 +03:00
Michael Crosby efb4ebcfe8 Merge pull request #6127 from unclejack/reuse_timestamp
reuse timestamp, don't call time.Now() 3 times
2014-05-30 17:21:19 -07:00
unclejack b64209f7b6 reuse timestamp, don't call time.Now() 3 times
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-31 03:03:51 +03:00
Michael Crosby 8ac516094e Merge pull request #5947 from glebfm/fix-socket-race-condition
Fix race condition during socket creation
2014-05-30 15:07:06 -07:00
unclejack b46449cb33 Merge pull request #6123 from vieux/no_number_default
no default default number in names
2014-05-31 00:55:11 +03:00
unclejack 0f40cedc23 Merge pull request #6124 from vieux/standadize_debug_iptables
use stderr to debug iptables
2014-05-31 00:46:28 +03:00
unclejack c90faa47b6 Merge pull request #6065 from vieux/fix_race_waitlock
update container's state after we close the waitLock
2014-05-31 00:38:35 +03:00
Victor Vieux 5708aa62f3 use stderr to debug iptables
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-30 19:39:42 +00:00
Timothy Hobbs 608702b980 Refactor device handling code
We now have one place that keeps track of (most) devices that are allowed and created within the container.  That place is pkg/libcontainer/devices/devices.go

This fixes several inconsistencies between which devices were created in the lxc backend and the native backend.  It also fixes inconsistencies between wich devices were created and which were allowed.  For example, /dev/full was being created but it was not allowed within the cgroup.  It also declares the file modes and permissions of the default devices, rather than copying them from the host.  This is in line with docker's philosphy of not being host dependent.

Docker-DCO-1.1-Signed-off-by: Timothy Hobbs <timothyhobbs@seznam.cz> (github: https://github.com/timthelion)
2014-05-30 19:21:29 +00:00
Victor Vieux e70a5ab014 no default default number in names
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-30 19:08:21 +00:00
Michael Crosby 251de0887b Merge pull request #6121 from unclejack/refactor_daemon_restore
optimize restore & use Getenv less in daemon.go
2014-05-30 11:37:12 -07:00
unclejack 0fd0deb75d optimize restore & use Getenv less in daemon.go
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-30 21:20:27 +03:00
Matt Heon ba1c0f4c81 Documented special case of container memory limits in manpages
If a container is started with a memory limit of 0, no memory limit is applied.
To prevent this from causing confusion, make interaction explicit in docs.

Docker-DCO-1.1-Signed-off-by: Matthew Heon <mheon@redhat.com> (github: mheon)
2014-05-30 13:55:13 -04:00
Michael Crosby 9e58a77e26 Merge pull request #6118 from LK4D4/fix_race_conditions
Fix race conditions
2014-05-30 10:43:56 -07:00
Alexandr Morozov adb639117b Atomically increment sequence in pkg/netlink
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-30 16:08:29 +04:00
Alexandr Morozov af17b01ad7 Goroutine-safe daemon.containers
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-30 16:08:06 +04:00
Alexandr Morozov 64bd6a6a53 Fix race in native driver on activeContainers usage
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-30 14:16:00 +04:00
Alexandr Morozov 4e5f6951f2 Fix races in set/get currentInterfaces in networkdriver
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-30 14:16:00 +04:00
Alexandr Morozov ce9e9ff4a1 Fix race on shutting down
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-30 14:15:46 +04:00
Alexandr Morozov c4990ab999 Fix races on TagStore accessing
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-30 13:36:14 +04:00
unclejack 838d6a9e9b Merge pull request #5855 from philips/remove-dead-code
chore(graph/graph): remove dead code
2014-05-30 04:51:12 +03:00
SvenDowideit 23dd221e52 sync the initial docs from b2d
Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-30 11:16:38 +10:00
Michael Crosby 107cd5dd5f Merge pull request #6109 from unclejack/block_aufs_on_top_of_btrfs
block aufs on incompatible file systems
2014-05-29 16:58:30 -07:00
unclejack e8a87120d4 block aufs on incompatible file systems
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-30 02:32:27 +03:00
Michael Crosby a86975f29b Merge pull request #6048 from crquan/use-timeout-conn
Use Timeout Conn wrapper to set read deadline for downloading layer
2014-05-29 15:23:21 -07:00
Michael Crosby 887ab77922 Merge pull request #6096 from vishh/stats_strongtype
Replaced fs.Stats() interface with fs.GetStats() which returns a cgroups.Stats strong type
2014-05-29 15:23:01 -07:00
Vishnu Kannan 72e6e5ff7e Added a new method cgroups.GetStats() which will return a cgroups.Stats object which will contain all the available cgroup Stats.
Remove old Stats interface in libcontainers cgroups package.
Changed Stats to use unit64 instead of int64 to prevent integer overflow issues.
Updated unit tests.

Docker-DCO-1.1-Signed-off-by: Vishnu Kannan <vishnuk@google.com> (github: vishh)
2014-05-29 20:16:49 +00:00
Michael Crosby 189c600b3b Merge pull request #6105 from gdm85/master
Do not consider iptables' output an error in case of xtables lock
2014-05-29 11:06:25 -07:00
Michael Crosby e2935f9c16 Merge pull request #6106 from discordianfish/improve-error-btrfs-on-non-btrfs
Add ErrPrerequisites to improve misleading errors
2014-05-29 11:00:13 -07:00
Victor Marmol f65f1660c0 Merge pull request #6098 from crosbymichael/retry-mount
Handle EBUSY on remount
2014-05-29 09:18:54 -07:00
Johannes 'fish' Ziemke 75754e69f6 Add ErrPrerequisites to improve misleading errors
There are two cases where we can't use a graphdriver:

1) the graphdriver itself isn't supported by the system
2) the graphdriver is supported by some configuration/prerequisites are
missing

This introduces a new error for the 2) case and uses it when trying to
run docker with btrfs backend on a non-btrfs filesystem.

Docker-DCO-1.1-Signed-off-by: Johannes 'fish' Ziemke <github@freigeist.org> (github: discordianfish)
2014-05-29 17:09:55 +02:00
Giuseppe Mazzotta 5e3b643ce6 * do not consider iptables' output an error in case of xtables lock
Docker-DCO-1.1-Signed-off-by: Giuseppe Mazzotta <gdm85@users.noreply.github.com> (github: gdm85)
2014-05-29 15:57:29 +02:00
Sven Dowideit 8622641b58 Merge pull request #6100 from SvenDowideit/pr_out_docs_installation_google_update_to_use_container_vm_images
docs/installation/google: update to use container-vm images
2014-05-29 11:41:43 +10:00
Johan Euphrosine 456ba11d89 docs/installation/google: update to use container-vm images
Docker-DCO-1.1-Signed-off-by: Johan Euphrosine <proppy@google.com> (github: proppy)

rebased by

Docker-DCO-1.1-Signed-off-by: Sven Dowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-29 11:36:28 +10:00
Sven Dowideit b9619bf957 Merge pull request #6099 from SvenDowideit/pr_out_updated_documentation_for_mentioning_images_vs_containers
updated documentation for mentioning images vs containers
2014-05-29 11:23:53 +10:00
Robin Speekenbrink a0505edc9a updated documentation for mentioning images vs containers
Docker-DCO-1.1-Signed-off-by: Robin Speekenbrink <robin@kingsquare.nl> (github: fruitl00p)

rebased by

Docker-DCO-1.1-Signed-off-by: Sven Dowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-29 11:20:41 +10:00
Michael Crosby 5d04b9deaf Handle EBUSY on remount
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-28 18:10:50 -07:00
Victor Vieux b204b97c9a Merge pull request #6083 from bernerdschaefer/nsinit-drop-capabilities-after-changing-user
SETUID/SETGID not required for changing user
2014-05-28 17:29:17 -07:00
unclejack 341a2b85be Merge pull request #5402 from SvenDowideit/dont-update-authors
Don't add yourself to the AUTHORs file - its generated
2014-05-29 02:00:29 +03:00
unclejack 745445ebb7 Merge pull request #6067 from SvenDowideit/go-fmt-s
gofmt does not result in the same thing as go fmt -s, which is what travis uses
2014-05-29 01:58:03 +03:00
unclejack a8131c7354 Merge pull request #6090 from alexlarsson/fix-console
Fix /dev/console
2014-05-29 01:09:33 +03:00
unclejack 180c2a6785 add support for COPY to docker build
This adds a COPY command to docker build which works like ADD, but is
only for local files and it doesn't extract files.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-29 00:07:52 +03:00
Alexander Larsson 7f5cd76824 libcontainer: Don't create a device node on /dev/console to bind mount on
There is no need for this, the device node by itself doesn't work, since
its not on a devpts fs, and we can just a regular file to bind mount over.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-28 21:07:40 +02:00
Alexander Larsson 35d08bdd01 Revert "Remove the bind mount for dev/console which override the mknod/label"
This reverts commit ae85dd5458.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-28 21:07:27 +02:00
unclejack 2330be2adc Merge pull request #6076 from LK4D4/remove_collections_package
Remove collections package
2014-05-28 21:32:27 +03:00
Victor Marmol 5e2af07137 Merge pull request #5868 from jhspaybar/5749-libcontainerroutes
libcontainer support for arbitrary route table entries
2014-05-28 10:50:56 -07:00
William Thurston bf7f360dca Fixes #5749
libcontainer support for arbitrary route table entries

Docker-DCO-1.1-Signed-off-by: William Thurston <me@williamthurston.com> (github: jhspaybar)
2014-05-28 17:42:02 +00:00
FLGMwt 50417f6abc Change 'Uploading context' wording
Docker-DCO-1.1-Signed-off-by: Ryan Stelly <ryan.stelly@live.com> (github: FLGMwt)
2014-05-28 11:59:29 -05:00
Bernerd Schaefer 0563453b91 SETUID/SETGID not required for changing user
It is no longer necessary to pass "SETUID" or "SETGID" capabilities to
the container when a "user" is specified in the config.

Docker-DCO-1.1-Signed-off-by: Bernerd Schaefer <bj.schaefer@gmail.com> (github: bernerdschaefer)
2014-05-28 16:41:48 +02:00
Bernerd Schaefer fd58524f81 Add system.SetKeepCaps and system.ClearKeepCaps
Docker-DCO-1.1-Signed-off-by: Bernerd Schaefer <bj.schaefer@gmail.com> (github: bernerdschaefer)
2014-05-28 16:40:36 +02:00
Sven Dowideit f00c2e5007 Merge pull request #5957 from cgwalters/manpage-typo
docker-run.1: Fix typo 'priviledged'
2014-05-28 22:43:46 +10:00
Alexandr Morozov 31f0a61a3d Remove collections package
It doesn't needed anymore аfter port and ip allocators refactoring
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4dmath@gmail.com> (github: LK4D4)
2014-05-28 13:59:45 +04:00
Derek 02f4ae6c56 Use Timeout Conn wrapper to set read deadline for downloading layer
Docker-DCO-1.1-Signed-off-by: Derek <crq@kernel.org> (github: crquan)
2014-05-27 22:50:04 -07:00
Sven Dowideit 77ae37a383 Merge pull request #6068 from jpetazzo/update-wording-private-self-hosted-registries
Replace 'private' with 'self-hosted' in a couple of places
2014-05-28 14:39:05 +10:00
Victor Vieux e56cf35c29 Merge pull request #6069 from crosbymichael/dm-errors
Return devmapper errors with additional text
2014-05-27 19:15:59 -07:00
Michael Crosby 6964012382 Return devmapper errors with additional text
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-27 19:03:57 -07:00
Jérôme Petazzoni b64535166f Replace 'private' with 'self-hosted' in a couple of places
Docker-DCO-1.1-Signed-off-by: Jérôme Petazzoni <jerome@docker.com> (github: jpetazzo)
2014-05-27 18:57:25 -07:00
SvenDowideit 147ad3480d gofmt does not result in the same thing as go fmt -s file.go, which is what travis tests for
Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-28 11:11:47 +10:00
O.S. Tezer d53eb2f67c Merge pull request #6053 from jamtur01/produpdate
Removed Docker production warnings
2014-05-27 18:00:49 -07:00
Victor Vieux 1de52caedc update container's state after we close the waitLock
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-28 00:18:51 +00:00
Sven Dowideit 58a0e8af8a Merge pull request #6044 from zdover23/27May_dockerfile_2
27 may dockerfile 2
2014-05-28 07:58:02 +10:00
Sven Dowideit b76302e5d9 Merge pull request #6040 from SvenDowideit/installer-docs
Rewrite the mac and windows docs to use the boot2docker installers.
2014-05-28 07:57:43 +10:00
SvenDowideit 5f5d66ade6 Rewrite the mac and windows docs to use the boot2docker installers.
Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-28 07:56:15 +10:00
Sven Dowideit 080d8a7a83 Merge pull request #6005 from jimperrin/master
Updated documentation to include CentOS installation instructions.
2014-05-28 07:55:03 +10:00
unclejack 70c6d806c8 Merge pull request #6063 from vieux/improve_number_detection
improve numeric only id detection
2014-05-28 00:36:57 +03:00
Victor Vieux b8a89ba963 Merge pull request #6062 from crosbymichael/update-waits
Update wait calls to call Wait on Command
2014-05-27 14:17:19 -07:00
Victor Vieux f2baa364a1 improve numeric only id detection
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-27 21:03:12 +00:00
Michael Crosby 5310e8575f Update lxc to use cmd.Wait()
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-27 13:52:05 -07:00
Michael Crosby b9de22e828 Update wait calls to call Wait on Command
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-27 13:38:24 -07:00
Victor Vieux a605379927 Merge pull request #6059 from erikh/fix-tests
Fix race condition in CLI tests: diff was not acquiring a container lock
2014-05-27 13:19:32 -07:00
unclejack 0d9a5ce6dd Merge pull request #6025 from crosbymichael/concurrent-names
Improve name generation on concurrent requests
2014-05-27 23:18:19 +03:00
Michael Crosby 88cefe20b4 Merge pull request #6030 from erikh/5971-fix-zombies
Add Wait() calls in the appropriate spots
2014-05-27 12:53:17 -07:00
Erik Hollensbe 8b77e0183e Fix race condition in CLI tests: diff was not acquiring a container lock
Docker-DCO-1.1-Signed-off-by: Erik Hollensbe <github@hollensbe.org> (github: erikh)
2014-05-27 12:28:31 -07:00
Erik Hollensbe b01c3283fa libcontainer/nsinit: remove Wait call from Exec and Kill from Attach in tty_term.go
Docker-DCO-1.1-Signed-off-by: Erik Hollensbe <github@hollensbe.org> (github: erikh)
2014-05-27 12:26:56 -07:00
Erik Hollensbe 92e41a02ce Add Wait() calls in the appropriate spots
Docker-DCO-1.1-Signed-off-by: Erik Hollensbe <github@hollensbe.org> (github: erikh)
2014-05-27 12:26:56 -07:00
Michael Crosby bdb5aa4c27 Merge pull request #5937 from LK4D4/ipallocator_refactoring
Ip allocator refactoring
2014-05-27 12:22:36 -07:00
Victor Vieux b904d0af56 Merge pull request #6051 from LK4D4/move_some_build_tests_to_integration_cli
Move some build tests to integration cli
2014-05-27 12:11:42 -07:00
Victor Vieux 754797bba7 Merge pull request #6000 from cyphar/5619-fix-unsafe-path-resolution
Properly handle paths with symlink path components
2014-05-27 12:06:19 -07:00
Alexandr Morozov ef94ac7d2f Ip allocator refactoring
We don't need ordered set anymore, also some cleanings and simple
benchmark.
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-27 22:56:26 +04:00
William Henry a14a88e53d Fixed some typos in docker-build.
Docker-DCO-1.1-Signed-off-by: William Henry <whenry@redhat.com> (github: ipbabble)

 Changes to be committed:
	modified:   contrib/man/md/docker-build.1.md
2014-05-27 12:05:48 -06:00
William Henry 2cb184ade5 Some changes to semantics around tagging, detaching.
Tried to clarify some of the semantics regarding --tag and renaming etc.
Added some text clarifying how to detach from a running containers.
Fixed a bug that reference /varlog instead of /dev/log.

Docker-DCO-1.1-Signed-off-by: William Henry <whenry@redhat.com> (github: ipbabble)

Changes to be committed:
	modified:   docker-attach.1.md
	modified:   docker-build.1.md
	modified:   docker-run.1.md
	modified:   docker-tag.1.md
2014-05-27 11:56:11 -06:00
Michael Crosby a22622b022 Merge pull request #5983 from vieux/add_tianon_mailmap_authors
add Tianon as maintainer of AUTHORS and .mailmap
2014-05-27 10:43:17 -07:00
Zac Dover 2bbf87504b fixed spacing, removed extraneous line
Docker-DCO-1.1-Signed-off-by: Zac Dover <zdover@redhat.com> (github: zdover23)
2014-05-28 03:32:52 +10:00
Tianon Gravi b2cf28306d Merge pull request #6029 from jeffreybolle/master
Update upstart script to ensure networking is up before starting docker
2014-05-27 11:23:54 -06:00
James Turnbull 838e69e11e Removed Docker production warnings
* Removed warnings.
* Removed inconsistent Community installation sections.
* Fixed all installation page descriptions.
* Removed old .inc files.

Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-27 13:08:19 -04:00
William Henry ca49da41ee Merge branch 'master' of https://github.com/dotcloud/docker 2014-05-27 09:11:12 -06:00
Jim Perrin b7e88250f4 Update documentation per suggestions from ostezer
Docker-DCO-1.1-Signed-off-by: Jim Perrin <jperrin@centos.org> (github: jimperrin)
2014-05-27 08:59:30 -05:00
James Turnbull cf47b46036 Merge pull request #6046 from thomasleveil/patch-1
Update ubuntulinux.md - bash completion
2014-05-27 05:39:39 -07:00
Thomas LEVEIL 250229605f Update ubuntulinux.md and debian.md
to add bash completion for the docker command
2014-05-27 14:17:01 +02:00
Alexandr Morozov b25a9b7138 Move entrypoint build test to integration-cli
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-27 08:36:38 +04:00
LK4D4 81d1641139 Move expose build test to integration-cli
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-27 08:00:35 +04:00
LK4D4 c58991f31a Move cmd build test to integration-cli
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-27 08:00:35 +04:00
LK4D4 b05be686ec Move env build test to integration-cli
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-27 08:00:35 +04:00
LK4D4 40630ce4b6 Move relative workdir build test to integration-cli
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-27 08:00:35 +04:00
LK4D4 360fb3d4ea Move user build test to integration-cli
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-27 08:00:34 +04:00
LK4D4 3dd4c5f499 Move maintainer build test to integration-cli
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-27 08:00:34 +04:00
LK4D4 11f7f0bf9b Move volume build test to integration-cli
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-27 08:00:34 +04:00
O.S. Tezer e5ba9f6443 Merge pull request #6042 from dcorking/patch-1
Update fedora.md
2014-05-26 12:27:45 -07:00
David Corking b26bebac31 Update fedora.md
The renamed package 'wmdocker' did not make it into Fedora 20. It is still available for rawhide (will be Fedora 21.)
2014-05-26 12:29:27 +01:00
Sven Dowideit 4a2ca56ea0 Merge pull request #6039 from zdover23/26May_dockerfile
Dockerfile.5.md stub
2014-05-26 16:46:12 +10:00
Zac Dover 41a913b390 Dockerfile.5.md stub
Properly formatted DCO:
Docker-DCO-1.1-Signed-off-by: Zac Dover <zdover@redhat.com> (github: zdover23)

Docker-DCO-1.1-Signed-off-by: Zac Dover <zdover@redhat.com> (github: )
2014-05-26 15:57:20 +10:00
Tianon Gravi bddcbc1b15 Merge pull request #6037 from tianon/fix-debootstrap-typo
Fix embarrassing debootstrap typo :)
2014-05-25 22:37:59 -06:00
Tianon Gravi dd3319636d Fix embarrassing debootstrap typo :)
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-25 21:59:43 -06:00
Jeffrey Bolle c52889db27 Ensure networking is up before starting docker
This resolves a problem that I have been having where docker starts before networking is up. See issue #5944 for more details.

Docker-DCO-1.1-Signed-off-by: Jeffrey Bolle <jeffreybolle@gmail.com> (github: jeffreybolle)
2014-05-24 19:03:11 +01:00
James Turnbull 7393eafde7 Merge pull request #6001 from jamtur01/introfixes
Some fixes to the Intro pages
2014-05-25 00:21:40 +10:00
James Turnbull 955bfda99a Some fixes to the Intro pages
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-25 00:20:59 +10:00
Jim Perrin 1cb2570a27 Correct appearance of shell output
Docker-DCO-1.1-Signed-off-by: Jim Perrin <jperrin@centos.org> (github: jimperrin)
2014-05-24 08:33:41 -05:00
Sven Dowideit 22777fbc1e Merge pull request #6006 from therealprologic/master
Updated installation docs for CRUX as a Docker Host.
2014-05-24 16:08:14 +10:00
Sven Dowideit d81912e323 Merge pull request #6016 from ronsmits/patch-1
Update using_supervisord.md
2014-05-24 16:07:32 +10:00
Sven Dowideit b820b9778f Merge pull request #6023 from proppy/patch-4
docs/reference/builder: fix USER doc
2014-05-24 13:41:34 +10:00
cyphar ff24a32876 integration-cli: cp: added symlink-related tests
This patch adds cli integration tests for #5619, which are tests
to ensure that symlinks are kept relative to the container rootfs
(even when a path component).

Docker-DCO-1.1-Signed-off-by: Aleksa Sarai <cyphar@cyphar.com> (github: cyphar)
2014-05-24 11:01:20 +10:00
cyphar 328d2cba11 daemon: container: properly handle paths with symlink path components
This patch fixes the incorrect handling of paths which contain a
symlink as a path component when copying data from a container.
Essentially, this patch changes the container.Copy() method to
first "resolve" the resource by resolving all of symlinks encountered
in the path relative to the container's rootfs (using pkg/symlink).

Docker-DCO-1.1-Signed-off-by: Aleksa Sarai <cyphar@cyphar.com> (github: cyphar)
2014-05-24 11:01:19 +10:00
Michael Crosby 6ec86cb6e5 Improve name generation on concurrent requests
Fixes #2586

This fixes a few races where the name generator asks if a name is free
but another container takes the name before it can be reserved.  This
solves this by generating the name and setting it.  If the set fails
with a non unique error then we try again.
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-23 17:51:16 -07:00
Johan Euphrosine e89cb9a5e0 docs/reference/builder: fix USER doc
Docker-DCO-1.1-Signed-off-by: Johan Euphrosine <proppy@google.com> (github: proppy)
2014-05-23 17:20:15 -07:00
Guillaume J. Charmes baed5d7433 Merge pull request #6020 from vieux/update_test_remove_nc_bind_mount
now busybox as nc
2014-05-23 14:48:24 -07:00
Michael Crosby 3b4b0a901d Merge pull request #6018 from vishh/stats_strongtype
Strong type all stats exported by libcontainer
2014-05-23 14:35:14 -07:00
Michael Crosby f3bd86376e Merge pull request #6022 from crosbymichael/iptables-wait
Add check for iptables xlock support
2014-05-23 14:25:51 -07:00
Michael Crosby 034babf175 Add check for iptables xlock support
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-23 14:18:50 -07:00
Victor Vieux 0be44d1a0a now busybox as nc
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-23 21:03:52 +00:00
Tibor Vass 7e38d6a358 Merge pull request #6019 from tianon/update-dockerfile-busybox
Add specific branch of @jpetazzo's busybox to clone from so we can assume nice things about it
2014-05-23 13:46:48 -07:00
Vishnu Kannan 321b457044 Added stats.go which provides strong types for all stats that will be exported by libcontainer. This commit only introduces the strong type.
Docker-DCO-1.1-Signed-off-by: Vishnu Kannan <vishnuk@google.com> (github: vishh)
2014-05-23 20:42:43 +00:00
Tianon Gravi 886d3c9396 Simplify "adduser" in the Dockerfile to use the more scripting-friendly "useradd"
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-23 14:32:43 -06:00
Tianon Gravi efa79a09a9 Add specific branch of jpetazzo's busybox to clone from so we can assume nice things about it
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-23 14:32:43 -06:00
Tibor Vass cfded04ac2 Merge pull request #6015 from crosbymichael/update-test-flags
Update integration tests with --net flag
2014-05-23 13:31:47 -07:00
Tibor Vass 174ab74755 Merge pull request #5998 from crosbymichael/iptables-wait
Add wait flag to iptables
2014-05-23 13:31:35 -07:00
Michael Crosby 2487237937 Update ip test to parse new output
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-23 13:22:01 -07:00
Ron Smits 9429fe6f49 Update using_supervisord.md
changed the version of ubuntu to 13.04. In latest sshd will not work.
2014-05-23 21:15:44 +02:00
Victor Vieux 0874b85627 Merge pull request #6008 from rhatdan/umount
We need to umount /var/lib/docker when the daemon exits.
2014-05-23 12:01:38 -07:00
Victor Vieux 494cbbec44 Merge pull request #5993 from LK4D4/fix_race_on_stop_#5154
Add synchronization between jobs and server closing
2014-05-23 11:31:10 -07:00
Michael Crosby 581e8e8918 Update integration tests with --net flag
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-23 11:31:01 -07:00
James Turnbull 973fae9d75 Merge pull request #5952 from nathanleclaire/patch-2
Fix a few issues.
2014-05-24 03:00:37 +10:00
Tianon Gravi 3c430e2599 Merge pull request #6012 from tianon/check-config-cgroup-subsystems
Add more required cgroup subsystems to check-config.sh
2014-05-23 10:43:16 -06:00
Nathan LeClaire 1e74cd0e01 Fix a few issues.
-single-dash => --double-dash for a few flags, and also there was a missing "B", as in "Container B".

Docker-DCO-1.1-Signed-off-by: Nathan LeClaire <nathan.leclaire@gmail.com> (github: nathanleclaire)
2014-05-23 09:36:27 -07:00
Tianon Gravi 0c0ffb7773 Add more required cgroup subsystems to check-config.sh (specifically, SCHED for cpu and CPUACCT for cpuacct)
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-23 10:27:09 -06:00
Jim Perrin 2a35e41a9c Squashed for pull request
Updated documentation to include CentOS installation instructions.
Extraneous line removal, description change

Docker-DCO-1.1-Signed-off-by: Jim Perrin <jperrin@centos.org> (github: jimperrin)
2014-05-23 10:46:01 -05:00
Dan Walsh a3ff8a98f7 We need to umount /var/lib/docker when the daemon exits.
Currently we are leaving it bind mounted on stop.

Docker-DCO-1.1-Signed-off-by: Dan Walsh <dwalsh@redhat.com> (github: rhatdan)
2014-05-23 11:43:46 -04:00
James Mills c0529e5fc1 Updated installation docs for CRUX as a Docker Host.
Docker-DCO-1.1-Signed-off-by: James Mills <prologic@shortcircuit.net.au> (github: therealprologic)
2014-05-24 01:18:00 +10:00
James Turnbull a6577708ab Merge pull request #5949 from nathanleclaire/patch-1
Change single-dash flag usage to double-dash
2014-05-23 13:43:23 +10:00
Michael Crosby b315c380f4 Add wait flag to iptables
Fixes #1573
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-23 01:24:58 +00:00
Tibor Vass be622a78da Merge pull request #5997 from crosbymichael/add-hang
Fix add hang when dest is .
2014-05-22 17:48:51 -07:00
Michael Crosby 04e357d09f Merge pull request #25 from tiborvass/3960-add-file-hangs
adding test for hanging ADD src .
2014-05-22 17:38:11 -07:00
Tibor Vass 1ce5457d57 adding test for hanging ADD src .
Docker-DCO-1.1-Signed-off-by: Tibor Vass <teabee89@gmail.com> (github: tiborvass)
2014-05-22 17:28:04 -07:00
Michael Crosby c8ada301ba Fix add hang when dest is .
Fixes #3960
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-23 00:26:38 +00:00
Michael Crosby 3d78c49aab Merge pull request #5995 from vieux/recur_nodes
Add device nodes recursively
2014-05-22 16:35:27 -07:00
Victor Vieux b6c65df093 update test
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-22 22:50:41 +00:00
Victor Vieux 0abb52c7a9 add recursive device nodes
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-22 22:29:13 +00:00
Michael Crosby 24a58a66e3 Merge pull request #5994 from vmarmol/libcontainers-stats
Make all cgroup stats output int64s instead of float64.
2014-05-22 14:24:14 -07:00
Victor Marmol 4a33a757d5 Make all cgroup stats output int64s instead of float64.
Docker-DCO-1.1-Signed-off-by: Victor Marmol <vmarmol@google.com> (github: vmarmol)
2014-05-22 20:53:36 +00:00
Alexandr Morozov f92d68a6ce Add synchronization between jobs and server closing
Fixes #5154
Daemon waiting 15 seconds for finishing server jobs before shutdown. In
this time it doesn't accept new jobs. After this time, it shutdown
despite running jobs.
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-22 22:54:49 +04:00
Michael Crosby db1a3551a3 Merge pull request #5839 from unclejack/improve_build_rm
add --force-rm to clean up after a failed build
2014-05-22 10:54:05 -07:00
Tianon Gravi 8006f6583d Merge pull request #5967 from tianon/consolidated-mkimage
Add new consolidated mkimage scripts
2014-05-22 11:27:08 -06:00
unclejack 32c1e3005f Merge pull request #5990 from alexlarsson/no-discard-mkfs
devmapper: Don't enable discard during mkfs.ext4
2014-05-22 19:38:21 +03:00
Alexander Larsson 42708181b1 devmapper: Don't enable discard during mkfs.ext4
There is no reason to do discard durink mkfs, as the filesystem
is on a newly allocated device anyway. Discard is a slow operation,
so this may help initial startup a bit, especially if you use a larger
thin pool.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-22 09:23:52 +02:00
James Turnbull 7470ebf30b Merge pull request #5746 from ImJasonH/patch-1
Update google.md
2014-05-22 13:43:40 +10:00
Victor Vieux 55d41c3e21 Merge pull request #5976 from crosbymichael/getpids
Move get pid into cgroup implementation
2014-05-21 19:09:50 -07:00
Victor Vieux 5a0a03e394 Merge pull request #5922 from crosbymichael/host-dev-priv
Mount /dev in tmpfs for privileged containers
2014-05-21 18:56:24 -07:00
James Turnbull 2fec15cff1 Merge pull request #5982 from jamtur01/bitbucket
Added BitBucket support to documentation
2014-05-22 11:35:17 +10:00
Nathan LeClaire e7babb28cc Change single-dash flag usage to double-dash
The single-dash long-form flag usage is deprecated.

Docker-DCO-1.1-Signed-off-by: Nathan LeClaire <nathanleclaire@gmail.com> (github: nathanleclaire)
2014-05-21 17:07:04 -07:00
James Turnbull cd44cd26cb Merge pull request #5959 from tpng/patch-1
Add instructions for persisting data
2014-05-22 09:55:05 +10:00
Sven Dowideit a0260c1181 Merge pull request #5897 from jamtur01/newintro
Rewrite of the Introduction documentation
2014-05-21 16:53:03 -07:00
Michael Crosby 4bb4bf634a Merge pull request #5859 from philips/append-etc-hosts-not-bind
fix(daemon): prepend host /etc/hosts instead of bind mounting
2014-05-21 15:57:59 -07:00
Victor Vieux fe1d297b9d Merge pull request #5895 from tianon/update-authors
Update AUTHORS and .mailmap
2014-05-21 22:57:05 +00:00
Victor Vieux cedd76cb52 Merge pull request #5981 from philips/bump-go-systemd
vendor: bump github.com/coreos/go-systemd to v2
2014-05-21 15:52:35 -07:00
Victor Vieux b1464b2daa add Tianon as maintainer of AUTHORS and .mailmap
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-21 22:48:31 +00:00
James Turnbull 62468a3eba Added BitBucket support to documentation
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-22 08:35:10 +10:00
Brandon Philips df5585b273 vendor: bump github.com/coreos/go-systemd to v2
There are a couple of bugfixes since this was last bumped. Pull them in.

Docker-DCO-1.1-Signed-off-by: Brandon Philips <brandon.philips@coreos.com> (github: philips)
2014-05-21 15:31:04 -07:00
Brandon Philips 61ac745d7a integration-cli: fix spelling error in test
Docker-DCO-1.1-Signed-off-by: Brandon Philips <brandon.philips@coreos.com> (github: philips)
2014-05-21 15:20:29 -07:00
Brandon Philips 5579bec47b integration-cli: tests for /etc/hosts and net=host
Some basic tests to make sure this is acting correctly on machines.

Docker-DCO-1.1-Signed-off-by: Brandon Philips <brandon.philips@coreos.com> (github: philips)
2014-05-21 15:20:29 -07:00
Michael Crosby 718e522c71 Merge pull request #5979 from rhatdan/terminal
This check was backwards.
2014-05-21 14:51:13 -07:00
Dan Walsh 26b4a4920a This check was backwards.
If I do a

docker events  > /tmp/out

I do not want the control characters getting written to the file.
The check should check the output file not the input file.

Docker-DCO-1.1-Signed-off-by: Dan Walsh <dwalsh@redhat.com> (github: rhatdan)
2014-05-21 17:37:40 -04:00
Michael Crosby 811d93326b Move get pid into cgroup implementation
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-21 21:14:07 +00:00
James Turnbull 0056884090 Rewrite of the Introduction documentation
1. Re-aligns the introduction with the new product positioning.
2. Cleanup of some issues with language and formatting.
3. Makes the introduction leaner and meaner.
4. Responds to feedback from product.

Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-22 07:11:43 +10:00
Victor Vieux b7f9e683c3 Merge pull request #5958 from erikh/5738-docker_port_in_use
docker port in use
2014-05-21 21:06:27 +00:00
Victor Vieux 457cb86cf2 Merge pull request #5975 from tianon/revert-run-tmpfs
Revert "Always mount a /run tmpfs in the container"
2014-05-21 13:46:48 -07:00
Tianon Gravi 8e967fe802 Revert "Always mount a /run tmpfs in the container"
This reverts commit 905795ece6.

Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-21 14:28:19 -06:00
Michael Crosby 2ff4f71528 Merge pull request #5953 from tiborvass/remove-chmod-755
remove chmod 755: fixes #5941
2014-05-21 12:27:06 -07:00
Tibor Vass bb431a7190 for perms checking tests, create files from Go and remove them from git
Docker-DCO-1.1-Signed-off-by: Tibor Vass <teabee89@gmail.com> (github: tiborvass)
2014-05-21 11:35:12 -07:00
Tibor Vass 56a53c72d2 remove chmod 755: fixes #5941
Docker-DCO-1.1-Signed-off-by: Tibor Vass <teabee89@gmail.com> (github: tiborvass)
2014-05-21 11:35:08 -07:00
James Turnbull 29a192ca4d Merge pull request #5965 from cpuguy83/5963-fix-single-dash-in-working-with-volumes
Fixes some docs issues with using single-dash arguments where they should be double
2014-05-22 03:31:46 +10:00
Victor Vieux 286a17ff19 Merge pull request #5955 from anandkumarpatel/5923-fix-leaking-go-routine
5923-fix-leaking-go-routine
2014-05-21 10:11:56 -07:00
Erik Hollensbe 91ba337931 bridge: retry any proxy set up failures until we get a successful port, defeats a race condition
Docker-DCO-1.1-Signed-off-by: Erik Hollensbe <github@hollensbe.org> (github: erikh)
2014-05-21 09:14:53 -07:00
Erik Hollensbe f0489ce3a9 portallocator: rewrite to simplify, removes race condition
Docker-DCO-1.1-Signed-off-by: Erik Hollensbe <github@hollensbe.org> (github: erikh)
2014-05-21 09:14:53 -07:00
Tianon Gravi 51f707cf9d Add deprecation warnings to the mkimage scripts that are have consolidated mkimage implementations
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-21 09:58:11 -06:00
Tianon Gravi 7e42505083 Add new consolidated mkimage scripts
These new scripts are streamlined such that, for example, "contrib/mkimage/debootstrap" is _only_ responsible for filling a directory with the results of running debootstrap, and it can accept any arbitrary arguments.

Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-21 09:47:36 -06:00
Brian Goff 6d9e64b27b Fixes some docs issues with using single-dash arguments where they should be double
I found a bunch of issues where we have "-<opt>" instead of "--<opt>".
Also a couple of other issues, like "-notrunc", which is now "--no-trunc"
Fixes #5963

Docker-DCO-1.1-Signed-off-by: Brian Goff <cpuguy83@gmail.com> (github: cpuguy83)
2014-05-21 09:35:22 -04:00
unclejack a94a87778c Merge pull request #5954 from vieux/fix_cp_error
Ignore lchown error on docker cp
2014-05-21 15:55:57 +03:00
Colin Walters cfe7211a22 man/docker-run.1: Fix typo 'priviledged', add missing closing brackets
Docker-DCO-1.1-Signed-off-by: Colin Walters <walters@verbum.org> (github: cgwalters)
2014-05-21 07:00:38 -04:00
tpng accea18920 Add instructions for persisting data
Docker-DCO-1.1-Signed-off-by: Benny Ng <benny.tpng@gmail.com> (github: tpng)
2014-05-21 16:45:41 +08:00
James Turnbull f37339c2db Merge pull request #5765 from jamtur01/contribf
Added to small patch exemption to cover Markdown/syntax errors
2014-05-21 17:13:03 +10:00
AnandkumarPatel 789a8f2616 use buffered channel so goroutine does not get blocked on done <- true when a timeout occurs.
Docker-DCO-1.1-Signed-off-by: Anandkumar Patel <anandkumarpatel@gmail.com> (github: anandkumarpatel)
2014-05-20 21:33:40 -07:00
Michael Crosby f042c3c157 Update code post codereview
Add specific types for Required and Optional DeviceNodes
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-21 00:40:41 +00:00
Victor Vieux 9e51b7abae ignore lchown error on docker cp
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-21 00:23:17 +00:00
Victor Vieux 5eef0a28cb add test
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-21 00:10:07 +00:00
Michael Crosby ed5892ed4e Update documentation for container struct in libcontainer
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-20 23:34:46 +00:00
Michael Crosby 34c05c58c8 Mount /dev in tmpfs for privileged containers
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-20 22:51:24 +00:00
Michael Crosby 83e9dc7200 Merge pull request #5755 from vieux/move_inspect_daemon
move inspect from server to daemon
2014-05-20 15:50:38 -07:00
Tianon Gravi 1f219672fa Merge pull request #5950 from tianon/update-deps
Update gorilla/mux, gorilla/context, and kr/pty deps
2014-05-20 16:16:21 -06:00
Michael Crosby 1310243d48 Merge pull request #5946 from crosbymichael/lxc-host-net
Fix network mode for lxc 1.0
2014-05-20 14:51:33 -07:00
Victor Vieux 240fad4974 Merge pull request #5951 from vieux/pr5919
Fix remote add cache
2014-05-20 14:33:05 -07:00
Victor Vieux 03a109e446 add test
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-20 21:31:28 +00:00
Guillaume J. Charmes bcfe2ceffb Remove the mtime for temp file. Prevent false negative cache
Docker-DCO-1.1-Signed-off-by: Guillaume J. Charmes <guillaume@charmes.net> (github: creack)
2014-05-20 21:31:28 +00:00
Brice Jaglin 9810da853b force the read of the tarSum so that sums actually get computed
Docker-DCO-1.1-Signed-off-by: Brice Jaglin <bjaglin@teads.tv> (github: bjaglin)
2014-05-20 21:31:28 +00:00
Michael Crosby 70d35b9d39 Merge pull request #5918 from crosbymichael/volumes-commit
Do not commit host bind mounts into image
2014-05-20 14:28:49 -07:00
Tianon Gravi d98af1236c Update gorilla/mux, gorilla/context, and kr/pty deps
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-20 15:23:50 -06:00
Michael Crosby e2d79bec3a Update runconfig unit test for changes
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-20 21:11:33 +00:00
Tianon Gravi a16cb394fa Merge pull request #5945 from crosbymichael/lxc-version-bump
Update LXC requirement to 1.0
2014-05-20 14:54:24 -06:00
Michael Crosby 94f01184be Update lxc requirement to 1.0
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-20 20:45:22 +00:00
Michael Crosby f915988a5d Merge pull request #5916 from rhatdan/crash
docker '' causes a golang crash.
2014-05-20 13:42:48 -07:00
Victor Vieux 603e00a3a7 move inspect from server to daemon
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-20 19:36:15 +00:00
Gleb Fotengauer-Malinovskiy 24c73ce2d3 Fix race condition during socket creation
Docker-DCO-1.1-Signed-off-by: Gleb Fotengauer-Malinovskiy <glebfm@altlinux.org> (github: glebfm)
2014-05-20 23:22:06 +04:00
Michael Crosby 0f27894094 Fix network mode for lxc 1.0
Fixes #5692

This change requires lxc 1.0+ to work and breaks lxc versions less than
1.0 for host networking.  We think that this is a find tradeoff by
bumping docker to only support lxc 1.0
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-20 19:10:23 +00:00
Dan Walsh bc081a03d8 docker '' causes a golang crash.
This patch fixes the problem.

Docker-DCO-1.1-Signed-off-by: Daniel Walsh <dwalsh@redhat.com> (github: rhatdan)
2014-05-20 14:04:27 -04:00
Michael Crosby e5202de918 Merge pull request #5930 from alexlarsson/add-cap-setfcap
native driver: Don't drop SETFCAP
2014-05-20 10:29:37 -07:00
Michael Crosby f3edb7c0e8 Merge pull request #5929 from alexlarsson/systemd-cgroup-allow-mknod
cgroups: Allow mknod for any device in systemd cgroup backend
2014-05-20 10:13:13 -07:00
Tianon Gravi 47aa9f5a0e Merge pull request #5844 from stpierre/init-script-wait-for-pidfile
Wait for pidfile to exist when starting with RedHat init script
2014-05-20 10:25:03 -06:00
Alexander Larsson fcf2e9a910 native driver: Add required capabilities
We need SETFCAP to be able to mark files as having caps, which is
heavily used by fedora.
See https://github.com/dotcloud/docker/issues/5928

We also need SETPCAP, for instance systemd needs this to set caps
on its childen.

Both of these are safe in the sense that they can never ever
result in a process with a capability not in the bounding set of the
container.

We also add NET_BIND_SERVICE caps, to be able to bind to ports lower
than 1024.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-20 11:31:39 +02:00
Alexander Larsson 6029504350 cgroups: Allow mknod for any device in systemd cgroup backend
Without this any container startup fails:
2014/05/20 09:20:36 setup mount namespace copy additional dev nodes mknod fuse operation not permitted

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-20 09:29:32 +02:00
Brandon Philips 000a37fe9d fix(daemon): prepend host /etc/hosts instead of bind mounting
systemd systems do not require a /etc/hosts file exists since an nss
module is shipped that creates localhost implicitly. So, mounting
/etc/hosts can fail on these sorts of systems, as was reported on CoreOS
in issue #5812.

Instead of trying to bind mount just copy the hosts entries onto the
containers private /etc/hosts.

Docker-DCO-1.1-Signed-off-by: Brandon Philips <brandon.philips@coreos.com> (github: philips)
2014-05-19 22:32:51 -07:00
Sven Dowideit bfe72c6189 Merge pull request #5920 from jmccrohan/master
docs: cli.md; document docker version output
2014-05-19 21:41:05 -07:00
Jonathan McCrohan 4052a81830 docs: cli.md; document docker version output
Commit 3cec63d56f changed the output of
docker version. Updating docs accordingly.

Docker-DCO-1.1-Signed-off-by: Jonathan McCrohan <jmccrohan@gmail.com> (github: jmccrohan)
2014-05-20 02:00:31 +01:00
Sven Dowideit 5511be7512 Merge pull request #5905 from jamtur01/faq
Adding the FAQ back
2014-05-20 10:17:01 +10:00
Michael Crosby d31c37fceb Add test for commiting container with bind mount
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-19 22:57:29 +00:00
Tianon Gravi 05c3325a02 Merge pull request #5890 from LK4D4/fix_tee_hanging_on_test_fails_#5672
Fix hanging on tests failing
2014-05-19 16:37:19 -06:00
Michael Crosby d535d98100 Don't save bind mounts in image
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-19 22:18:37 +00:00
Michael Crosby e454be7567 Move init volumes code to func
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-19 22:04:51 +00:00
Guillaume J. Charmes b5d0bdc9e8 Merge pull request #5915 from vieux/start_remove_link
starts by removing the name from the graph db
2014-05-19 14:38:55 -07:00
Guillaume J. Charmes 2d1c245631 Merge pull request #5913 from crosbymichael/fuse
Make sure dev/fuse is created in container
2014-05-19 14:31:32 -07:00
Victor Vieux c12dd77090 starts by removing the name from the graph db
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-19 21:22:33 +00:00
Michael Crosby b466672a4c Merge pull request #5914 from unclejack/fix_unprivileged_user_creation
add unprivilegeduser via the Dockerfile
2014-05-19 14:08:06 -07:00
unclejack 599cb12bb8 add unprivilegeduser via the Dockerfile
This changes the test TestBuildWithInaccessibleFilesInContext to not
add the user 'unprivilegeduser' and add it via the Dockerfile instead.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-19 23:55:28 +03:00
Michael Crosby a87bcefb8b Make sure dev/fuse is created in container
Fixes #5849

If the host system does not have fuse enabled in the kernel config we
will ignore the is not exist errors when trying to copy the device node
from the host system into the container.
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-19 20:46:59 +00:00
Guillaume J. Charmes 5fee3774d4 Merge pull request #5911 from LK4D4/chmod_on_add_#3979
Make chmod on ADDed files
2014-05-19 13:35:55 -07:00
unclejack a691fcb277 integcli: add tests for build --rm
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-19 23:10:54 +03:00
unclejack b60d647172 default to deleting images via the remote api
This makes the remote API version 1.12 and newer default to
automatically deleting intermediate containers when the build has
succeedeed.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-19 23:10:54 +03:00
unclejack 69dcf767fd integcli: test container removal for failed builds
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-19 23:10:54 +03:00
unclejack 77f5425260 integcli: add getContainerCount utility function
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-19 23:10:54 +03:00
unclejack 667e2bd4ea bump remote api to 1.12 & add --force-rm to build
This adds a `--force-rm` flag to docker build which makes the Docker
daemon clean up all containers, even when the build has failed.

This new flag requires that we bump the remote API, so we also bump the
remote API version.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-19 23:10:49 +03:00
Alexandr Morozov 620c8c7253 Make chmod on ADDed files
Fixes #3979
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-20 00:09:39 +04:00
unclejack e39299ca1d docs: add back build to remote api v1.10 & v1.11
This adds back the rm query parameter to the remote api docs for api
v1.10 and v1.11.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-19 22:46:19 +03:00
Victor Marmol 30bd2bbc83 Merge pull request #5903 from alexlarsson/writable-proc
Make /proc writable, but not /proc/sys and /proc/sysrq-trigger
2014-05-19 12:21:15 -07:00
Victor Vieux d74b6095c9 Merge pull request #5861 from crquan/fix-user-agent-trailing-space
Remove Trailing Whitespace in User-Agent
2014-05-19 12:08:53 -07:00
Michael Crosby b2275ee8df Merge pull request #5869 from vbatts/vbatts-vendor-capabilities
gocapability: upstream fix for unsporrted caps
2014-05-19 12:08:12 -07:00
Alexander Larsson 68493e2f7f Make /proc writable, but not /proc/sys and /proc/sysrq-trigger
Some applications want to write to /proc. For instance:

docker run -it centos groupadd foo

Gives: groupadd: failure while writing changes to /etc/group

And strace reveals why:

open("/proc/self/task/13/attr/fscreate", O_RDWR) = -1 EROFS (Read-only file system)

I've looked at what other systems do, and systemd-nspawn makes /proc read-write
and /proc/sys readonly, while lxc allows "proc:mixed" which does the same,
plus it makes /proc/sysrq-trigger also readonly.

The later seems like a prudent idea, so we follows lxc proc:mixed.
Additionally we make /proc/irq and /proc/bus, as these seem to let
you control various hardware things.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-19 20:46:05 +02:00
Victor Vieux ccf29f0b94 Merge pull request #5879 from timthelion/docs-sysinit-readme
Add README.md file for sysinit directory
2014-05-19 11:26:22 -07:00
Victor Vieux 7fb5ce44a0 Merge pull request #5875 from timthelion/docs-archive-readme
Add readme file for the archive directory
2014-05-19 11:25:55 -07:00
Victor Marmol cb7680b9b9 Merge pull request #5792 from bernerdschaefer/nsinit-supports-pdeathsig
Add PDEATHSIG support to nsinit library
2014-05-19 11:13:23 -07:00
Victor Vieux 8674ec7f75 Merge pull request #5871 from unclejack/speed_up_buildsixtysteps
integcli: speed up TestBuildSixtySteps
2014-05-19 11:02:36 -07:00
Victor Vieux 40c8fc668c Merge pull request #5876 from timthelion/docs-daemon-readme
Add readme for daemon directory
2014-05-19 10:51:17 -07:00
Victor Vieux 5690e03553 Merge pull request #5877 from timthelion/docs-daemon-config-readme
Add readme for the daemonconfig directory
2014-05-19 10:50:51 -07:00
Victor Vieux 3fa96c55e3 Merge pull request #5874 from timthelion/docs-api-readme
Add readme file for the api directory
2014-05-19 10:50:00 -07:00
Victor Vieux 8eef1be29e Merge pull request #5782 from unclejack/fix_5270 2014-05-19 10:36:10 -07:00
Michael Crosby b50c9a185c Merge pull request #5670 from jmccrohan/master
client: Remove docker new version check
2014-05-19 10:26:58 -07:00
Michael Crosby 15a94fdcbb Merge pull request #5864 from shykes/pr_out_make_cristian_maintainer_of_integration_cli_he_wrote_it
Make Cristian maintainer of integration-cli (he wrote it)
2014-05-19 10:26:18 -07:00
Michael Crosby a3a26cd349 Merge pull request #5907 from vmarmol/caps
Don't drop CAP_FOWNER in the container.
2014-05-19 10:06:29 -07:00
Michael Crosby 265de539ff Merge pull request #5865 from crosbymichael/add-all-caps
Add the rest of the caps so that they are retained in privilged mode
2014-05-19 09:56:55 -07:00
Michael Crosby 66e7cf24b7 Merge pull request #5880 from LK4D4/fix_uid_overflow_#5647
Check uid ranges
2014-05-19 09:55:36 -07:00
Victor Marmol 0abad3ae22 Don't drop CAP_FOWNER in the container. Also sorts the list of allowed
capabilities.

Docker-DCO-1.1-Signed-off-by: Victor Marmol <vmarmol@google.com> (github: vmarmol)
2014-05-19 16:52:39 +00:00
Michael Crosby e1c7abe890 Add the rest of the caps so that they are retained in privilged mode
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-19 16:43:31 +00:00
James Turnbull 702442b586 Adding the FAQ back
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-19 10:07:57 -04:00
James Turnbull e5422e212f Merge pull request #5884 from brandon-rhodes/master
Expand the Advanced Networking page to all options
2014-05-19 03:35:48 -04:00
James Turnbull 59e37784ee Merge pull request #5834 from waitingkuo/patch-1
Update ubuntulinux.md
2014-05-19 03:34:14 -04:00
Alexandr Morozov 7cd1e48230 Fix hanging on tests failing
Tee hanging when `go test` exit with non-zero code.
Fixes #5672
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-19 08:26:05 +04:00
Brandon Rhodes ca9c35cdf7 Expand the Advanced Networking page to all options
For issue #5658 this rewrite of the networking page explains what every
single option (unless I missed one!) does both for the Docker server and
also the Docker client when submitting a "docker run".  I somehow
thought that, when I was done, there would be a lot more about setting
up topologies like I did for my Foundations of Python Network
Programming network of Docker containers.  More about making routers and
firewalls that NAT and so forth.  But, at least for this draft, I think
that setting up subnets and setting up point-to-point links is most of
what Docker users will need unless they are doing something exotic.

We can always expand later.

Docker-DCO-1.1-Signed-off-by: Brandon Rhodes <brandon@rhodesmill.org> (github: brandon-rhodes)

Improve networking.md documentation per @jamtur01

Besides catching some typos and awkward sentences, @jamtur01 had several
more thorough changes to suggest:

 *  He illustrated the correct way to delimit "Note" paragraphs in
     Docker documentation.

 *  He helped point out where I was presuming an Ubuntu host running
    Docker, so that I could re-word those sections to be specific that
    the advice only applied to Ubuntu (these mostly related to how to
    set server-wide options and restart the server).

 *  He is happy to have "Ethernet" capitalized even where Linux
    documentation would render it with an ugly lower-case first letter.

Docker-DCO-1.1-Signed-off-by: Brandon Rhodes <brandon@rhodesmill.org> (github: brandon-rhodes)

Improve networking.md per ideas from @SvenDowideit

A response to a bit of further discussion of pull request #5884.
2014-05-18 21:58:39 -04:00
James Turnbull 3b96078856 Merge pull request #5883 from jamtur01/debianinst
Added basic Debian installation page
2014-05-18 19:02:48 -04:00
James Turnbull f8974b5cfc Fixed sudo section to match Debian installation doc
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-18 18:55:58 -04:00
James Turnbull 88afc8992f Added basic Debian installation page
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-18 18:55:36 -04:00
James Turnbull 255eca7193 Merge pull request #5889 from financeCoding/couchdb_doc_nit
removing grep command that depends on perl, better for boot2docker
2014-05-18 23:29:29 +02:00
Tianon Gravi 89c4748b83 Update AUTHORS and .mailmap
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-18 13:26:13 -06:00
Adam Singer 7f9178c6d1 removing grep command that depends on perl, better for boot2docker
Docker-DCO-1.1-Signed-off-by: Adam Singer <financeCoding@gmail.com> (github: financeCoding)
2014-05-18 10:02:39 -07:00
Alexandr Morozov 72d1e40c4a Check uid ranges
Fixes #5647
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-18 20:49:08 +04:00
O.S. Tezer f65db7c565 Merge pull request #5882 from jamtur01/ubfix
Fixed description and keywords on Ubuntu installation docs
2014-05-18 17:26:29 +01:00
O.S. Tezer a2dffd72ef Merge pull request #5886 from financeCoding/mac_doc_nit
command line nit
2014-05-18 17:17:47 +01:00
Adam Singer 21e36ab36c command line nit 2014-05-17 22:04:41 -07:00
Jonathan McCrohan 3cec63d56f client: Rip out HTTP check from docker version
For background to this change please see:
https://github.com/dotcloud/docker/issues/4802
https://github.com/dotcloud/docker/pull/5670

Docker-DCO-1.1-Signed-off-by: Jonathan McCrohan <jmccrohan@gmail.com> (github: jmccrohan)
2014-05-18 02:22:22 +01:00
Jonathan McCrohan 82712ed67e client: rearrange docker version output
Rearrange docker version output so that server output matches client
output

Docker-DCO-1.1-Signed-off-by: Jonathan McCrohan <jmccrohan@gmail.com> (github: jmccrohan)
2014-05-18 01:49:58 +01:00
James Turnbull a52db7d880 Fixed description and keywords on Ubuntu installation docs
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-17 22:18:08 +02:00
Timothy f9728de7a3 Add README.md file for sysinit directory
Note, this text is just copied from sysinit.go

See Issue #5873

Docker-DCO-1.1-Signed-off-by: Timothy <timothyhobbs@seznam.cz> (github: https://github.com/timthelion)
2014-05-17 18:12:44 +00:00
Timothy 8dfd4b677b Add readme for the daemonconfig directory
See Issue #5873

Docker-DCO-1.1-Signed-off-by: Timothy <timothyhobbs@seznam.cz> (github: https://github.com/timthelion)
2014-05-17 18:01:47 +00:00
Timothy b175b4dd43 Add readme for daemon directory
See issue #5873

Docker-DCO-1.1-Signed-off-by: Timothy <timothyhobbs@seznam.cz> (github: https://github.com/timthelion)
2014-05-17 17:56:02 +00:00
Timothy 28fc387cf0 Add readme file for the archive directory
See issue #5873

Docker-DCO-1.1-Signed-off-by: Timothy <timothyhobbs@seznam.cz> (github: https://github.com/timthelion)
2014-05-17 17:48:07 +00:00
Timothy 3e2b97ef26 Add readme file for the api directory
See issue #5873

Docker-DCO-1.1-Signed-off-by: Timothy <timothyhobbs@seznam.cz> (github: https://github.com/timthelion)
2014-05-17 17:37:05 +00:00
unclejack 58c11ee0a8 integcli: speed up TestBuildSixtySteps
This improves the TestBuildSixtySteps test by switching from busybox to
scratch and simply adding a file.

This lowers the execution time of that test from 20 seconds to 5
seconds.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-17 17:25:56 +03:00
James Turnbull 9405a58e99 Merge pull request #5850 from rhatdan/fixmanpage
Add description of --volume read-only/read-write to docker-run man page.
2014-05-17 10:50:18 +02:00
O.S. Tezer 4dabca9046 Merge pull request #5832 from stevenburgess/documentation_cli_fix
Refactor b.tar -> busybox.tar
2014-05-17 09:42:20 +01:00
Vincent Batts 4bf03a0fac gocapability: upstream fix for unsporrted caps
@vmarmol has made the fix upstream for not failing if the capability
being dropped is returned as invalid from the syscall, which is the case
when the capability is not supported on the host.
This is a blocker presently for RHEL6.5 on CAP_SYSLOG. We have patched
around this in our RPM for the time being, but this is the proper fix.

See also https://github.com/dotcloud/docker/pull/5810

Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-05-17 03:51:02 -04:00
Solomon Hykes ab6027324d Make Cristian maintainer of integration-cli (he wrote it)
Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-05-16 19:03:08 -07:00
Michael Crosby cc84dd0967 Merge pull request #5863 from vieux/pr5862
Make libcontainer's CapabilitiesMask into a []string (Capabilities). + 	add support for CAP_FOWNER
2014-05-16 18:23:50 -07:00
Victor Vieux 9e7757a3d2 Merge pull request #5827 from vieux/fix_multiple_events_on_same_computer
allow 2 docker events to be opened on the same computer over unix socket
2014-05-16 18:17:46 -07:00
Victor Vieux a0070f0c17 add support for CAP_FOWNER
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-17 01:16:07 +00:00
Victor Marmol 92614928ce Make libcontainer's CapabilitiesMask into a []string (Capabilities).
Docker-DCO-1.1-Signed-off-by: Victor Marmol <vmarmol@google.com> (github: vmarmol)
2014-05-17 00:44:10 +00:00
Derek 42734394b0 Remove Trailing Whitespace in User-Agent
After removed, the User-Agent shows in log like this:

[debug] http.go:160 https://index.docker.io/v1/repositories/busybox/images --
HEADERS: map[User-Agent:[docker/0.11.1-dev go/go1.2.2 git-commit/8887e00-dirty kernel/3.14.3-n1 os/linux arch/amd64]]

The code also moved all validation work into validVersion,
to keep the main logic as clean.

Docker-DCO-1.1-Signed-off-by: Derek <crq@kernel.org> (github: crquan)
2014-05-16 17:15:04 -07:00
Victor Vieux 704e9f9ff0 Merge pull request #5761 from SvenDowideit/pr_out_tell_the_user_not_to_run_from_osx
tell the user not to run -d from OSX
2014-05-16 16:52:08 -07:00
unclejack f5b1afae74 add test for issue #5270
Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-17 02:38:29 +03:00
unclejack 33d5b38d62 use ValidateContextDirectory to validate context
This commit makes the Docker cli client use ValidateContextDirectory
before attempting to create a tarball out of the context. This ensures
we avoid errors such as "unexpected EOF" during the upload of the
context.

This check is done before uploading any data and can save time and
bandwidth for remote Docker daemons.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-17 02:38:29 +03:00
unclejack 1dedcd0d37 add ValidateContextDirectory to utils/utils.go
This commit adds a function which can be used to ensure all contents of
a directory can be accessed.

This function doesn't follow symlinks to check if they're pointing to
files which exist. Such symlinks can be useful later.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-17 02:38:28 +03:00
unclejack 46578a2359 integcli: resolve full path to docker binary
Setting dockerBinary to the full path of the Docker binary is a good
idea and this is now done in the test code.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-17 02:38:28 +03:00
Guillaume J. Charmes 6f651ec62b Merge pull request #5851 from proppy/archive-simple-detect
archive: simplify DetectCompression
2014-05-16 16:32:55 -07:00
Michael Crosby eb61a1f73a Merge pull request #5822 from creack/update_vendors
Update archive/tar vendored
2014-05-16 16:31:37 -07:00
Johan Euphrosine 4a3aefbb52 archive: add missing bytes package
Docker-DCO-1.1-Signed-off-by: Johan Euphrosine <proppy@google.com> (github: proppy)
2014-05-16 16:26:28 -07:00
Johan Euphrosine 67ce59a806 archive: fix panic if len(source) < len(m)
Docker-DCO-1.1-Signed-off-by: Johan Euphrosine <proppy@google.com> (github: proppy)
2014-05-16 16:25:56 -07:00
Johan Euphrosine 3118952e0a archive: simplify DetectCompression
Docker-DCO-1.1-Signed-off-by: Johan Euphrosine <proppy@google.com> (github: proppy)
2014-05-16 16:25:56 -07:00
Michael Crosby 58b49e60bb Merge pull request #4882 from vieux/fix_merge_issue_tty
Fix merge issue tty
2014-05-16 15:42:43 -07:00
Victor Vieux 2c8b63cb75 do not merge -i or -t options
Docker-DCO-1.1-Signed-off-by: Victor Vieux <victor.vieux@docker.com> (github: vieux)
2014-05-16 22:31:16 +00:00
Victor Vieux 366f084eb0 add test
Docker-DCO-1.1-Signed-off-by: Victor Vieux <victor.vieux@docker.com> (github: vieux)
2014-05-16 22:23:32 +00:00
Brandon Philips a42e451e10 chore(graph/graph): remove dead code
This code was introduced with commented out sections. Just delete them.

Original commit: 18fc707fdf

Docker-DCO-1.1-Signed-off-by: Brandon Philips <brandon.philips@coreos.com> (github: philips)
2014-05-16 14:26:14 -07:00
Victor Vieux 00ce30bc74 Merge pull request #5571 from vieux/fix_race_docker_daemon
Fix race docker daemon
2014-05-16 13:57:43 -07:00
Victor Vieux 1adc2fa533 Merge pull request #5853 from proppy/update-tarfilter-comment
archive: update TarFilter comments
2014-05-16 13:06:12 -07:00
Victor Vieux 4b587ac118 Merge pull request #5436 from hobofan/master
reapply Cmd value even if CmdAdd returns early
2014-05-16 12:45:23 -07:00
Johan Euphrosine 311614952e archive: update TarFilter comments
Docker-DCO-1.1-Signed-off-by: Johan Euphrosine <proppy@google.com> (github: proppy)
2014-05-16 12:44:42 -07:00
Dan Walsh 2e0d47e5c4 Add description of --volume read-only/read-write to docker-run man page.
Docker-DCO-1.1-Signed-off-by: Dan Walsh <dwalsh@redhat.com> (github: rhatdan)
2014-05-16 15:06:34 -04:00
Michael Crosby 62c3183fc8 Merge pull request #5833 from ActiveState/fix_nsinit_env_panic
fix panic when passing empty environment
2014-05-16 12:03:26 -07:00
Sridhar Ratnakumar d787f2731e fix panic when passing empty environment
Docker-DCO-1.1-Signed-off-by: Sridhar Ratnakumar <github@srid.name> (github: srid)
2014-05-16 11:55:34 -07:00
Victor Marmol 01d10d6f13 Merge pull request #5810 from vmarmol/drop-caps
Change libcontainer to drop all capabilities by default.
2014-05-16 11:51:41 -07:00
Michael Crosby 23062e55fd Merge pull request #5396 from lalyos/5370-infinite-loop-caused-by-symlink
5370 infinite loop caused by symlink
2014-05-16 11:28:37 -07:00
Victor Vieux 808d28033b Merge pull request #5841 from unclejack/run_pull_latest
run: pull only latest when no tag specified
2014-05-16 11:21:29 -07:00
Michael Crosby dbd6f10b3d Merge pull request #5842 from alexlarsson/dm-cleanup-tests
devicemapper: Remove non-general tests
2014-05-16 11:08:04 -07:00
Victor Vieux 2b3b03b596 Merge pull request #5837 from proppy/patch-2
server/buildfile: correct daemon.Create comments
2014-05-16 10:51:09 -07:00
Johan Euphrosine 8b31824ad6 server/buildfile: correct daemon.Create comments
s/and start it//

Docker-DCO-1.1-Signed-off-by: Johan Euphrosine <proppy@google.com> (github: proppy)
2014-05-16 10:50:02 -07:00
Victor Vieux 021ac36a70 Merge pull request #5838 from timthelion/patch-1
Typo in execdrivers.go five => give
2014-05-16 10:49:44 -07:00
Tianon Gravi 6dd4fa0b74 Merge pull request #5840 from unclejack/upstart_mounted
Update contrib/init/upstart to "start on local-filesystems"
2014-05-16 10:16:34 -06:00
unclejack ba0c829291 contrib/init/upstart: start on local-filesystems
This changes the upstart init script to start on `local-filesystems`.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-16 19:12:59 +03:00
Chris St. Pierre 1ee423bd5d added timeout when waiting for docker pidfile
Docker-DCO-1.1-Signed-off-by: Chris St. Pierre <chris.a.st.pierre@gmail.com> (github: stpierre)
2014-05-16 09:36:54 -04:00
Chris St. Pierre a598dba134 Wait for pidfile to exist when starting with Redhat init script
Fixes #5359

Docker-DCO-1.1-Signed-off-by: Chris St. Pierre <chris.a.st.pierre@gmail.com> (github: stpierre)
2014-05-16 09:08:59 -04:00
unclejack b8e338144e run: pull only latest when no tag specified
This makes Docker pull only the image tagged as latest when no tag has
been specified. This makes Docker pull only the image it'll run.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-16 15:23:47 +03:00
Alexander Larsson 39d244a593 devmapper: Remove sys.go mocking framework
This is no longer needed, as the tests are gone.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-16 14:12:56 +02:00
Bernerd Schaefer 6a1d76bc7b nsinit.DefaultCreateCommand sets Pdeathsig to SIGKILL
Docker-DCO-1.1-Signed-off-by: Bernerd Schaefer <bj.schaefer@gmail.com> (github: bernerdschaefer)
2014-05-16 13:48:41 +02:00
Bernerd Schaefer 00e1adfead nsinit.Init() restores parent death signal before exec
Docker-DCO-1.1-Signed-off-by: Bernerd Schaefer <bj.schaefer@gmail.com> (github: bernerdschaefer)
2014-05-16 13:48:41 +02:00
Timothy Hobbs 408f050d64 Typo in execdrivers.go five => give
Docker-DCO-1.1-Signed-off-by: Timothy Hobbs <timothyhobbs@seznam.cz> (github: timthelion)
2014-05-16 11:45:20 +00:00
O.S. Tezer c4fc591fbe Merge pull request #5828 from SvenDowideit/pr_out_update_the_cli_md_docs_with_the_output_of_the_docker_command
Update the cli.md docs with the output of the docker command
2014-05-16 12:10:55 +01:00
Alexander Larsson bd6fd25dfb devicemapper: Remove non-general tests
Now that we have the generic graphtest tests that actually tests
the driver we can remove the old mock-using tests. Almost all of
these tests were disabled anyway, and the four remaining ones
didn't really test much while at the same time being really
fragile and making the rest of the code more complex due to
the mocking setup.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-16 12:46:28 +02:00
waitingkuo 1b4915c1a8 Update ubuntulinux.md
In Ubuntu 14.04, we should use

    $ sudo service docker.io restart

instead of

    $ sudo service docker restart
2014-05-16 17:08:05 +08:00
Steven Burgess ddb99054bc Refactor b.tar -> busybox.tar
The file was saved as busybox.tar, but the ls commands named it b.tar.

Docker-DCO-1.1-Signed-off-by: Steven Burgess <steven.a.burgess@hotmail.com> (github: stevenburgess)
2014-05-15 22:33:28 -04:00
Victor Marmol 9d6875d19d Change libcontainer to drop all capabilities by default. Only keeps
those that were specified in the config. This commit also explicitly
adds a set of capabilities that we were silently not dropping and were
assumed by the tests.

Docker-DCO-1.1-Signed-off-by: Victor Marmol <vmarmol@google.com> (github: vmarmol)
2014-05-16 00:57:58 +00:00
Michael Crosby 08a7bcf5db Merge pull request #5497 from alexlarsson/generic-driver-tests
graphdriver: Add generic test framework for graph drivers
2014-05-15 17:21:09 -07:00
Guillaume J. Charmes 4353ee78e8 Merge pull request #5454 from kzys/utime-omit
UTIME_OMIT is only available on Linux
2014-05-15 17:18:37 -07:00
Solomon Hykes 87fc7130a4 Merge pull request #5461 from SvenDowideit/getmaintainers.sh 2014-05-15 17:15:40 -07:00
Sven Dowideit 6297f333bd Merge pull request #5802 from shykes/pr_out_add_licensing_information_to_readme
Add Licensing information to README
2014-05-16 10:08:46 +10:00
Sven Dowideit 8da44c9b2a Merge pull request #5732 from SvenDowideit/pr_out_the_last_remaining_raw_github_com_url_s
the last remaining raw.github.com URL's
2014-05-16 10:01:14 +10:00
SvenDowideit d35330bb7a Update the cli.md docs with the output of the docker command
Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-16 10:00:20 +10:00
Victor Vieux 81ffd6530a allow 2 docker events to be opened on the same computer over unix socket
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-15 23:27:36 +00:00
Guillaume J. Charmes 924aa984cf Merge pull request #5823 from proppy/archive-bufio
archive: use bufio for compression detection
2014-05-15 16:08:21 -07:00
lalyos ad35d522db Fixes 5370 infinite/maxLoopCount loop for relative symlinks
use path.IsAbs() instead of checking if first char is '/'

Docker-DCO-1.1-Signed-off-by: Lajos Papp <lajos.papp@sequenceiq.com> (github: lalyos)
2014-05-16 01:03:11 +02:00
lalyos b51c366bfc Defend against infinite loop when following symlinks
ideally it should never reach it, but there was already multiple issues with infinite loop
at following symlinks. this fixes hanging unit tests

Docker-DCO-1.1-Signed-off-by: Lajos Papp <lajos.papp@sequenceiq.com> (github: lalyos)
2014-05-16 00:47:20 +02:00
lalyos 8b77a5b7ae Adding test case for symlink causes infinit loop, reproduces: dotcloud#5370
normally symlinks are created as either
ln -s /path/existing /path/new-name
or
cd /path && ln -s ./existing new-name

but one can create it this way
cd /path && ln -s existing new-name

this drives FollowSymlinkInScope into infinite loop

Docker-DCO-1.1-Signed-off-by: Lajos Papp <lajos.papp@sequenceiq.com> (github: lalyos)
2014-05-16 00:47:20 +02:00
Guillaume J. Charmes aedcb76adc Add test for invalid tarball read
Docker-DCO-1.1-Signed-off-by: Guillaume J. Charmes <guillaume@charmes.net> (github: creack)
2014-05-15 15:45:40 -07:00
Guillaume J. Charmes 094926206b Update archive/tar vendored
Docker-DCO-1.1-Signed-off-by: Guillaume J. Charmes <guillaume@charmes.net> (github: creack)
2014-05-15 15:40:26 -07:00
James Turnbull e5a5808f51 Merge pull request #5820 from joelhandwell/patch-1
changed deprecated option styles to supported ones
2014-05-16 00:40:16 +02:00
Victor Vieux 041dbcb7be Merge pull request #5826 from tianon/add-cross-to-test
Add "cross" to the "test" Makefile target
2014-05-15 15:35:18 -07:00
Tianon Gravi 8ec01f9dc1 Add "cross" to the "test" Makefile target
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-15 16:21:17 -06:00
Johan Euphrosine 77c2b76e44 archive: use bufio for compression detection
Docker-DCO-1.1-Signed-off-by: Johan Euphrosine <proppy@google.com> (github: proppy)
2014-05-15 11:41:07 -07:00
Joel Handwell 10a320818e changed deprecated option styles to supported ones
-name and -rm to --name and --rm
2014-05-15 10:50:59 -04:00
Sven Dowideit 5dfe7c43d4 Merge pull request #5816 from unclejack/docs_run_cache
docs: explain when RUN cache gets invalidated
2014-05-15 19:54:15 +10:00
unclejack 239d1ef301 docs: explain when RUN cache gets invalidated
This adds a few lines to the RUN Dockerfile docs to explain how to
bypass the RUN caching and how ADD can also invalidate the cache for
these instructions.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-15 12:32:01 +03:00
Bernerd Schaefer 002aa8fc20 Add GetParentDeathSignal() to pkg/system
Docker-DCO-1.1-Signed-off-by: Bernerd Schaefer <bj.schaefer@gmail.com> (github: bernerdschaefer)
2014-05-15 10:17:44 +02:00
Sven Dowideit 8887e00630 Merge pull request #5762 from SvenDowideit/pr_out_add_some_more_specific_help_for_dockerfile_build_contexts
Add some more specific help for Dockerfile build contexts.
2014-05-15 11:03:22 +10:00
Solomon Hykes 8d213e91b3 Merge pull request #5706 from vieux/remove_add_string 2014-05-14 17:31:52 -07:00
James Turnbull db4d534b5b Merge pull request #5800 from jamtur01/paras
Updated a bunch of formatting in the docs/sources/use files
2014-05-15 01:58:14 +02:00
SvenDowideit ed703fc8be Add some more specific help for Dockerfile build contexts.
I've copy and pasted the error message so its googlable.

Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-15 09:41:05 +10:00
Sven Dowideit d9e3d88082 Merge pull request #5801 from jamtur01/fixups
Minor fixups to the Docs README
2014-05-15 09:38:52 +10:00
James Turnbull 2269472f3a Updated a bunch of formatting in the docs/sources/use files
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-15 01:26:36 +02:00
Michael Crosby 6ff6aad9b8 Merge pull request #5809 from crosbymichael/move-cgroups
Move cgroups package into libcontainer
2014-05-14 16:06:32 -07:00
Michael Crosby b22d10e3c5 Remove the cgroups maintainer file
We don't need this because it is covered by the libcontainer MAINTAINERS
file
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-14 16:01:45 -07:00
Michael Crosby b0578d0ac7 Merge pull request #5772 from unclejack/improve_commit_performance
process directories in parallel in ChangesDirs
2014-05-14 15:49:11 -07:00
Michael Crosby 3b7a19def6 Move cgroups package into libcontainer
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-14 15:21:44 -07:00
Tianon Gravi 8904dfe082 Merge pull request #5657 from tianon/sysvinit-should-start
Add "Should-Start" cgroupfs-mount and cgroup-lite to sysvinit-debian init script
2014-05-14 16:05:32 -06:00
Michael Crosby 76f09c2ecf Merge pull request #5807 from crosbymichael/dev-symlinks
Setup standard /dev symlinks
2014-05-14 14:17:33 -07:00
Victor Vieux 49deffe6fe Merge pull request #5806 from icecrime/5766-events_endless_newlines
Issue #5766 - Check channel closed state in /events loop
2014-05-14 14:09:26 -07:00
O.S. Tezer 343c24a5fa Merge pull request #5805 from joelhandwell/patch-2
changed deprecated -name option to --name
2014-05-14 22:09:17 +01:00
O.S. Tezer 79be9f8d7f Merge pull request #5804 from joelhandwell/patch-1
changed deprecated -rm option to --rm
2014-05-14 22:08:01 +01:00
unclejack 359f8aca29 process directories in parallel in ChangesDirs
This commit lowers the total time spent in ChangesDirs to half during a
commit.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-15 00:03:33 +03:00
Bernerd Schaefer 2bc34036b9 Setup standard /dev symlinks
After copying allowed device nodes, set up "/dev/fd", "/dev/stdin",
"/dev/stdout", and "/dev/stderr" symlinks.

Docker-DCO-1.1-Signed-off-by: Bernerd Schaefer <bj.schaefer@gmail.com> (github: bernerdschaefer)
[rebased by @crosbymichael]
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-14 13:59:13 -07:00
James Turnbull f664e9a7d9 Added to small patch exemption to cover Markdown/syntax errors in documentation
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-14 22:43:54 +02:00
Arnaud Porterie 8699f53e6a Check channel closed state in /events loop
When listener channel is closed, it becomes always available for
reading: select becomes an active loop which writes default-constructed
events (i.e: empty strings).

Fixes #5766.

Docker-DCO-1.1-Signed-off-by: Arnaud Porterie <arnaud.porterie@gmail.com> (github: icecrime)
2014-05-14 22:35:04 +02:00
Guillaume J. Charmes 17a1f470ae Merge pull request #5783 from LK4D4/fix_duplicate_ip_allocation_#5729
Fix duplicate ip allocation
2014-05-14 13:32:27 -07:00
Joel Handwell 2d622df835 changed deprecated -name option to --name
changed deprecated -name option to --name
2014-05-14 16:24:08 -04:00
Joel Handwell 137f4b326a changed deprecated -rm option to --rm
changed deprecated -rm option to --rm
2014-05-14 16:22:38 -04:00
Michael Crosby 4af465fccf Merge pull request #5720 from cyphar/5656-cp-absolute-paths
Ensure `docker cp` cannot traverse outside container rootfs
2014-05-14 11:46:11 -07:00
Victor Vieux d066ceaa4f Merge pull request #5752 from crosbymichael/fix-port-mapping-ps
Fix port mapping in ps display for public and private
2014-05-14 11:36:23 -07:00
Victor Vieux bc22c9948c Merge pull request #5756 from crosbymichael/move-units-to-pkg
Move duration and size to units pkg
2014-05-14 11:36:14 -07:00
Victor Vieux c78b390b6c Merge pull request #5780 from vbatts/vbatts-start_tarsum_test
tarsum: start a test for TarSum
2014-05-14 11:28:26 -07:00
James Turnbull a50a46048b Minor fixups to the Docs README
Docker-DCO-1.1-Signed-off-by: James Turnbull <james@lovedthanlost.net> (github: jamtur01)
2014-05-14 20:14:06 +02:00
James Turnbull d9b1c1976f Merge pull request #5787 from SvenDowideit/pr_out_add_a_mention_of_80_column_lines_and_reflow_the_document_to_hide_the_evidence
Add a mention of 80 column lines and reflow the document to hide the evidence.
2014-05-14 20:13:48 +02:00
Michael Crosby de406b69f3 Merge pull request #5798 from unclejack/fix_startup_speed
don't update Suffixarray for every Register during startup
2014-05-14 11:13:02 -07:00
Michael Crosby 432e42e715 Merge pull request #5791 from bernerdschaefer/nsinit-exec-forwards-signals
"nsinit exec ..." forwards signals to container
2014-05-14 11:05:27 -07:00
Victor Vieux 3bf1b562e3 Merge pull request #5781 from creack/remove_bind_console
Remove the bind mount for dev/console which override the mknod/label
2014-05-14 10:57:21 -07:00
unclejack fc82199d4f Merge pull request #5758 from crosbymichael/add-cpuset
Add cpuset cpus support for docker
2014-05-14 20:57:12 +03:00
Victor Vieux b561f0398a Merge pull request #5794 from unclejack/speed_up_ps
don't call sort for every add in daemon/history
2014-05-14 10:48:08 -07:00
Vincent Batts d153740d9c tarsum: adding benchmarks
to cover a couple of use-cases:
* 1mb file, using no compression
* 1mb file, using compression
* 1024 1k files, using no compression
* 1024 1k files, using compression

Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-05-14 11:48:17 -04:00
unclejack 5d5c89398c update Suffixarray only once during daemon startup
This commit makes the Docker daemon call UpdateSuffixarray only after
it finishes registering all containers.

This lowers the amount of time required for the Docker daemon to start
up.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-14 17:58:37 +03:00
unclejack 219b7ae8b5 add UpdateSuffixarray and refactor TruncIndex
This commit refactors TruncIndex to make it possible to add container
ids to the TruncIndex without updating the Suffixarray.

This is useful during the Docker daemon's startup when we don't want to
update the Suffixarray for every container we add.

Add continues to function like before.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-14 17:57:59 +03:00
unclejack e963179c79 don't call sort for every add in history
This moves the call to sort in daemon/history to a function to be
called explicitly when we're done adding elements to the list.

This speeds up `docker ps`.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-14 15:02:02 +03:00
Bernerd Schaefer 830c2d7fa3 "nsinit exec ..." forwards signals to container
Docker-DCO-1.1-Signed-off-by: Bernerd Schaefer <bj.schaefer@gmail.com> (github: bernerdschaefer)
2014-05-14 11:01:02 +02:00
James Turnbull cb0f2a2823 Merge pull request #5785 from SvenDowideit/pr_out_subject_and_object_containers_can_be_confused
Subject and object containers can be confused
2014-05-14 07:56:31 +02:00
Alexandr Morozov 80fca061e7 Change ip allocation logic
Now IP reuses only after all IPs from network was allocated
Fixes #5729

Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-14 06:43:10 +04:00
Alexandr Morozov 5128feb690 Refactoring collections/orderedintset and benchmarks for it
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-14 06:04:12 +04:00
Michael Crosby adbe3096e8 Add cpuset cpus support for docker
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-13 18:17:12 -07:00
Michael Crosby 4de6810be9 Merge pull request #5786 from crosbymichael/cpuset-copy
Copy parents cpus and mems for cpuset
2014-05-13 18:16:47 -07:00
cyphar 0fb507dc23 daemon: *: refactored container resource path generation
This patch is a preventative patch, it fixes possible future
vulnerabilities regarding unsantised paths. Due to several recent
vulnerabilities, wherein the docker daemon could be fooled into
accessing data from the host (rather than a container), this patch
was created to try and mitigate future possible vulnerabilities in
the same vein.

Docker-DCO-1.1-Signed-off-by: Aleksa Sarai <cyphar@cyphar.com> (github: cyphar)
2014-05-14 11:15:50 +10:00
cyphar 79ca77f3e8 integration-cli: cp: added tests for cp
This patch adds integration tests for the copying of resources
from a container, to ensure that regressions in the security of
resource copying can be easily discovered.

Docker-DCO-1.1-Signed-off-by: Aleksa Sarai <cyphar@cyphar.com> (github: cyphar)
2014-05-14 11:14:59 +10:00
cyphar bfc3a4192a daemon: container: ensure cp cannot traverse outside container rootfs
This patch fixes the bug that allowed cp to copy files outside of
the containers rootfs, by passing a relative path (such as
../../../../../../../../etc/shadow). This is fixed by first converting
the path to an absolute path (relative to /) and then appending it
to the container's rootfs before continuing.

Docker-DCO-1.1-Signed-off-by: Aleksa Sarai <cyphar@cyphar.com> (github: cyphar)
2014-05-14 11:14:59 +10:00
Michael Crosby 3de15bda7e Copy parents cpus and mems for cpuset
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-13 18:01:31 -07:00
SvenDowideit 195f3a3f42 Add a mention of 80 column lines and reflow the document to hide the
evidence.

Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-14 10:22:55 +10:00
SvenDowideit af891a6751 Subject and object containers can be confused
and `create` is now a loaded word for some readers

Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-14 07:43:41 +10:00
Vincent Batts cfd1227e91 tarsum: test gofmt
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-05-13 15:14:32 -04:00
Vincent Batts efa369a6ee tarsum: adding the layer for "scratch" image
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-05-13 15:08:48 -04:00
Guillaume J. Charmes ae85dd5458 Remove the bind mount for dev/console which override the mknod/label
Docker-DCO-1.1-Signed-off-by: Guillaume J. Charmes <guillaume@charmes.net> (github: creack)
2014-05-13 11:59:27 -07:00
Vincent Batts 461f801f83 tarsum: start a test for TarSum
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-05-13 14:57:31 -04:00
Victor Vieux f637eaca5d Merge pull request #5778 from crosbymichael/check-symlink
Ensure libcontainer follows correct symlink in scope
2014-05-13 11:28:00 -07:00
Michael Crosby ea7647099f Add MAINTAINERS file to symlink pkg
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-13 11:27:24 -07:00
Michael Crosby ca040b1a37 Update code to handle new path to Follow Symlink func
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-13 10:54:08 -07:00
Michael Crosby dcf81f95fd Move Follow symlink to pkg
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-13 10:35:57 -07:00
Alexander Larsson b7c7b851dc libcontainer: Ensure bind mount target files are inside rootfs
Before we create any files to bind-mount on, make sure they are
inside the container rootfs, handling for instance absolute symbolic
links inside the container.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-13 10:24:52 -07:00
Sven Dowideit 1d4caadfe2 Merge pull request #5769 from dankeder/master
Fix a typo in contrib/man/md/docker.1.md
2014-05-13 20:26:02 +10:00
Dan Keder 653328c6ce Fix a typo in contrib/man/md/docker.1.md 2014-05-13 11:17:42 +02:00
Sven Dowideit a70bffd382 Merge pull request #5745 from ojilles/patch-1
Cleaned up Network settings overview
2014-05-13 19:02:18 +10:00
unclejack f1eabe436a Merge pull request #5655 from alexlarsson/mount-run-dir
Always mount a /run tmpfs in the container
2014-05-13 11:51:14 +03:00
Solomon Hykes ee4fa10fb0 Merge pull request #5753 from creack/fix-beam-32bit
[beam] Use direct iota instead of binary shift
2014-05-13 00:21:32 -07:00
James Turnbull 5ac3c6c4ed Merge pull request #5757 from SvenDowideit/pr_out_was_checking_something_for_a_pr_and_noticed_some_quote_issues
Was checking something for a PR and noticed some quote issues
2014-05-13 06:37:12 +02:00
SvenDowideit fe445a2447 tell the user not to run from OSX
Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-13 12:51:00 +10:00
Michael Crosby 5b525feaed Merge pull request #5707 from vieux/fix_event_removal
fix event removal
2014-05-12 18:59:42 -07:00
Victor Vieux 937f8f2d81 move acceptconnections as builtin
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-13 01:53:38 +00:00
Victor Vieux f3736265fd make listen buffer optional
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-13 01:51:41 +00:00
Victor Vieux a0f799b6d4 Merge pull request #5759 from vieux/move_register_links_to_daemon
move RegisterLinks to daemon
2014-05-12 18:39:47 -07:00
Victor Vieux 100a92146e move RegisterLinks to daemon
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-13 00:54:46 +00:00
SvenDowideit b9c2d57d89 Was checking something for a PR and noticed some quote issues
Docker-DCO-1.1-Signed-off-by: SvenDowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-13 10:36:19 +10:00
Sven Dowideit 3f6ec6ebba Merge pull request #5747 from ojilles/patch-2
runmetrics update, linking to collectd plugin rather than just hinting
2014-05-13 10:21:34 +10:00
Michael Crosby d33b4655c4 Move duration and size to units pkg
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-12 17:05:07 -07:00
Michael Crosby d54dec4d8b Fix port mapping in ps display for public and private
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-12 15:26:23 -07:00
Guillaume J. Charmes 0d78799db7 Merge pull request #5751 from vieux/move_version
move version out of server
2014-05-12 15:25:40 -07:00
Victor Vieux 7894a70f8b move version out of server
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-12 21:38:20 +00:00
Guillaume J. Charmes bc77ec2b17 Merge pull request #5717 from philips/mkdir-in-daemon
fix(daemon): ensure the /var/lib/docker dir exists
2014-05-12 14:13:56 -07:00
Solomon Hykes 43e926cd9c Merge pull request #5750 from vieux/vieux_server_maintainer
add vieux as server's MAINTAINERS
2014-05-12 13:56:30 -07:00
Victor Vieux 1d15c0b491 add vieux as server's MAINTAINERS
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-12 20:51:45 +00:00
Michael Crosby aa1fc36a92 Merge pull request #5696 from cyphar/rev-ubuntu-dockerfile
Update Ubuntu release used for builds
2014-05-12 13:21:38 -07:00
Alexander Larsson 905795ece6 Always mount a /run tmpfs in the container
All modern distros set up /run to be a tmpfs, see for instance:
https://wiki.debian.org/ReleaseGoals/RunDirectory

Its a very useful place to store pid-files, sockets and other things
that only live at runtime and that should not be stored in the image.

This is also useful when running systemd inside a container, as it
will try to mount /run if not already mounted, which will fail for
non-privileged container.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-12 21:41:04 +02:00
Michael Crosby 30a40de205 Merge pull request #5748 from crosbymichael/libcontainer-bindmounts
libcontainer: Create dirs/files as needed for bind mounts
2014-05-12 12:27:18 -07:00
Michael Crosby cc678a7078 Remove newline char in error message
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-12 12:24:30 -07:00
Michael Crosby 243eaac8b6 Merge pull request #5724 from tutumcloud/5718-lxcbugfix
Fix bug on LXC container start. Fixes #5718
2014-05-12 12:14:57 -07:00
Michael Crosby 16fe76e770 Merge pull request #5712 from vishh/fix_typo
Correct a libcontainer Mount Namespace setup comment.
2014-05-12 12:07:19 -07:00
Vishnu Kannan fee1bbd79e Correct a comment in libcontainer Mount Namespace setup.
Docker-DCO-1.1-Signed-off-by: Vishnu Kannan <vishnuk@google.com> (github: vishh)
2014-05-12 19:01:36 +00:00
Jilles Oldenbeuving 79f234e049 runmetrics update, linking to collectd plugin rather than just hinting 2014-05-12 20:51:40 +02:00
Jason Hall 3e636446c3 Update google.md
Fix numbering resets after code sample blocks
2014-05-12 14:48:03 -04:00
Jilles Oldenbeuving feb42d3f97 Cleaned up Network settings overview 2014-05-12 20:35:18 +02:00
unclejack 4c06506994 Merge pull request #5729 from fabiofalci/link_unlink_test
Integration test for link and unlink containers
2014-05-12 21:06:03 +03:00
Victor Vieux 34b7e352eb Merge pull request #5741 from creack/remove_unused_file-2
Remove signal_freebsd (already in pkg/signal)
2014-05-12 10:47:45 -07:00
Thatcher 27a39bfd1a Merge pull request #5621 from ostezer/docs-add-pardot
Add code for Pardot implementation to docs
2014-05-12 09:18:12 -07:00
O.S. Tezer 0f88dbd744 Merge pull request #5731 from SvenDowideit/reduce-cache-max-age
reduce cache time to 1 hours so future docs releases update the cloudfront caches faster
2014-05-12 17:05:55 +01:00
O.S. Tezer 8700d68b19 Merge pull request #5730 from SvenDowideit/fix-hostname-for-redirects
Give the Redirect a HostName - filled in from the s3 bucket name.
2014-05-12 15:03:03 +01:00
Alexander Larsson 70ef53f25e libcontainer: Create dirs/files as needed for bind mounts
If you specify a bind mount in a place that doesn't have a file yet we
create that (and parent directories). This is needed because otherwise
you can't use volumes like e.g. /dev/log, as that gets covered by the
/dev tmpfs mounts.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-12 09:57:15 +02:00
Sven Dowideit c1108f5d60 the last remaining raw.github.com URL's
Docker-DCO-1.1-Signed-off-by: Sven Dowideit <SvenDowideit@home.org.au> (github: SvenDowideit)
2014-05-12 14:50:41 +10:00
Sven Dowideit 1dec8fd03c reduce cache time to 1 hours so future docs releases update the cloudfront caches faster
Docker-DCO-1.1-Signed-off-by: Sven Dowideit <SvenDowideit@fosiki.com> (github: SvenDowideit)
2014-05-12 11:31:44 +10:00
Sven Dowideit 0d59cc080a Give the Redirect a HostName - filled in from the s3 bucket name.
Docker-DCO-1.1-Signed-off-by: Sven Dowideit <SvenDowideit@fosiki.com> (github: SvenDowideit)
2014-05-12 10:31:27 +10:00
Fabio Falci 7cc27b2075 Integration test for link and unlink containers
Docker-DCO-1.1-Signed-off-by: Fabio Falci <fabiofalci@gmail.com> (github: fabiofalci)
2014-05-11 21:33:01 +01:00
Guillaume J. Charmes 62bfef59f7 Use int64 instead of int
Docker-DCO-1.1-Signed-off-by: Guillaume J. Charmes <guillaume@charmes.net> (github: creack)
2014-05-11 06:23:53 -07:00
Vincent Mayers a4a023580b Add Licensing information to README
Docker-DCO-1.1-Signed-off-by: Vincent Mayers vince.mayers@gmail.com (github: vincentmayers)
[solomon@docker.com: rephrased commit message and transcribed signoff]
Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-05-11 08:03:31 +00:00
Guillaume J. Charmes 3f8ffb461a Remove signal_freebsd (already in pkg/signal)
Docker-DCO-1.1-Signed-off-by: Guillaume J. Charmes <guillaume@charmes.net> (github: creack)
2014-05-11 01:03:12 -07:00
Fernando 752c57ae56 Fix bug on LXC container start. Fixes #5718
Docker-DCO-1.1-Signed-off-by: Fernando Mayo <fernando@tutum.co> (github: fermayo)
2014-05-10 18:05:02 -04:00
Solomon Hykes 5877ae2462 Merge pull request #5716 from shykes/pr_out_engine_receiver_and_sender_preserve_job_env 2014-05-09 19:47:32 -07:00
Solomon Hykes 4dc16ab546 Merge pull request #5432 from shykes/image-subsystem-1 2014-05-09 19:19:33 -07:00
Solomon Hykes ff4ef50470 Start separating the image subsystem
This is a first step towards moving all code related to local
manipulation of images into a cleanly separated subsystem,
accessible via a stable set of commands in the engine API.

`graph.TagStore` now implements `engine.Installer`. For now, it
is installed by `Server.InitServer`, along with all other Server
commands. However this will change in future patches.

`graph.TagStore.Install` registers the following commands:

* `image_set` creates a new image and stores it locally.
* `image_get` returns information about an image stored locally.
* `image_tag` assigns a new name and tag to an existing image.

These commands are a pre-requisite for moving 'push' and 'pull'
out of `Server`.

Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-05-09 19:15:26 -07:00
Solomon Hykes d6e666a87a Merge pull request #5635 from vbatts/vbatts-registry_maintainers 2014-05-09 19:07:25 -07:00
cyphar 46a47f0dcb Dockerfile: updated ubuntu release used for builds
This patch just revs the Ubuntu release used for builds from 13.10
to 14.04.

Docker-DCO-1.1-Signed-off-by: Aleksa Sarai <cyphar@cyphar.com> (github: cyphar)
2014-05-10 11:38:12 +10:00
Brandon Philips b4ccd7cbfb fix(daemon): ensure the /var/lib/docker dir exists
The btrfs driver attempts to stat the /var/lib/docker directory to
ensure it exists. If it doesn't exist then it bails with an unhelpful
log line:

```
2014/05/10 00:51:30 no such file or directory
```

In 0.10 the directory was created but quickly digging through the logs I
can't tell what sort of re-ordering of code caused this regression.

Docker-DCO-1.1-Signed-off-by: Brandon Philips <brandon.philips@coreos.com> (github: philips)
2014-05-09 18:09:26 -07:00
Solomon Hykes a1754c7e46 Engine: Receiver and Sender preserve Job.Env
When sending a new job to a `engine.Sender`, the corresponding
`engine.Receiver` will receive that job with its environment preserved.
Previously the job name, arguments and streams were preserved but the
env was lost.

Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-05-09 17:10:33 -07:00
Solomon Hykes c7978c9809 Engine: Env.MultiMap, Env.InitMultiMap: import/export to other formats
* `Env.MultiMap` returns the contents of an Env as `map[string][]string`
* `Env.InitMultiMap` initializes the contents of an Env from a `map[string][]string`

This makes it easier to import and export an Env to other formats
(specifically `beam/data` messages)

Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-05-09 17:06:32 -07:00
Solomon Hykes 2af030ab57 beam/data: Message.GetOne() returns the last value set at a key
This is a convenience for callers which are only interested in one value
per key. Similar to how HTTP headers allow multiple keys per value, but
are often used to store and retrieve only one value.

Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-05-09 17:01:27 -07:00
Solomon Hykes 9dc66f8822 Merge pull request #5713 from shykes/pr_out_engine_ensure_all_pipes_are_properly_closed_by_receiver_and_sender 2014-05-09 16:16:56 -07:00
Solomon Hykes d61190169d Engine: ensure all pipes are properly closed by Receiver and Sender
Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)

[michael@docker.com: fix stdin closing in engine.Job.Run]
[michael@docker.com: fix fd leak in engine.Receiver.Run]
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@docker.com> (github: crosbymichael)

Docker-Tested-By: Solomon Hykes <solomon@docker.com>
Docker-Tested-by: Michael Crosby <michael@docker.com>
2014-05-09 16:12:21 -07:00
Solomon Hykes a7e61a21c0 Merge pull request #5687 from shykes/pr_out_engine_fix_a_timeout_bug_in_sender_receiver 2014-05-09 15:50:53 -07:00
Solomon Hykes bf25951837 Engine: slightly more stress-testing of Receiver/Sender to reproduce the hang problem
Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-05-09 15:49:44 -07:00
Solomon Hykes dfdc03b061 Engine: fix a timeout bug in Sender/Receiver
Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-05-09 15:49:44 -07:00
Solomon Hykes 0aeff69e59 Fix stdin handling in engine.Sender and engine.Receiver
This introduces a superficial change to the Beam API:

* `beam.SendPipe` is renamed to the more accurate `beam.SendRPipe`
* `beam.SendWPipe` is introduced as a mirror to `SendRPipe`

There is no other change in the beam API.

Docker-DCO-1.1-Signed-off-by: Solomon Hykes <solomon@docker.com> (github: shykes)
2014-05-09 15:39:55 -07:00
Victor Vieux 670564d070 fix event removal
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-09 22:12:54 +00:00
Michael Crosby da0686481c Merge pull request #5709 from vieux/conflict_h_net
returns an error when using -h and --net
2014-05-09 14:53:46 -07:00
Victor Vieux 2899195540 returns an error when using -h and --net
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-09 21:42:22 +00:00
Jérôme Petazzoni 29f1efc940 Merge pull request #5463 from tianon/hack-dind-cgroup
Update hack/dind to mount cgroups on "/cgroup" instead of "/sys/fs/cgroup" for better compatibility
2014-05-09 13:54:52 -07:00
Victor Vieux 0f6fe22833 Merge pull request #5686 from crosbymichael/cast-to-closer
Cast Input and Output to io.Closer
2014-05-09 11:17:53 -07:00
Victor Vieux e304e3a662 remove addString and replace Tail
Docker-DCO-1.1-Signed-off-by: Victor Vieux <vieux@docker.com> (github: vieux)
2014-05-09 18:09:59 +00:00
Aaron Huslage 5dfe0945c3 Fixing pre-commit-msg URL
URL referred to raw.github.com instead of raw.githubusercontent.com. Fixed.

Docker-DCO-1.1-Signed-off-by: Aaron Huslage <huslage@gmail.com> (github: huslage)
2014-05-09 13:56:20 -04:00
O.S. Tezer d4c774de97 Merge pull request #5697 from mshytikov/fix/doc
Added value format description for VOLUME instruction in builder doc
2014-05-09 17:14:27 +01:00
Max Shytikov b5a37127aa Added value format description for VOLUME instruction
In the documentation was not mentioned explicitly that VOLUME value
shoud be a valid JSON array. Because of this I spent time to discovering
the problem with my image where I put `VOLUME ['/data']` (with single quotes).
The `['/data']` mount point was parsed and mounted whole as a string without
any errors and warnings.

Docker-DCO-1.1-Signed-off-by: Max Shytikov <mshytikov@gmail.com> (github: mshytikov)
2014-05-09 16:37:55 +02:00
Alexander Larsson 55cd7dd7f9 grapdriver: Skip tests on non-supported backends
For now this means the btrfs backend is skipped when run
inside make test. You can however run it manually if you want.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-09 14:48:39 +02:00
Alexander Larsson 4bdb8c03fc graphdriver: Fail initialization if supported but got error
If a graphdriver fails initialization due to ErrNotSupported we ignore
that and keep trying the next. But if some driver has a different
error (for instance if you specified an unknown option for it) we fail
the daemon startup, printing the error, rather than falling back to an
unexected driver (typically vfs) which may not match what you have run
earlier.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-09 14:48:39 +02:00
Alexander Larsson 27744062aa graphdriver: Add generic test framework for graph drivers
This adds daemon/graphdriver/graphtest/graphtest which has a few
generic tests for all graph drivers, and then uses these
from the btrs, devicemapper and vfs backends.

I've not yet added the aufs backend, because i can't test that here
atm. It should work though.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-09 14:48:39 +02:00
Alexander Larsson 84f19a09ac vfs graphdriver: Make root dir mode 755
This matches the other backends.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-09 14:48:38 +02:00
Alexander Larsson 47c79870ea devmapper: Properly restore mocked functions after test
Currently the tests that mocks or denies functions leave this state
around for the next test. This is no good if we want to actually
test the devicemapper code in later tests.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-09 14:48:38 +02:00
James Turnbull 79dc316a86 Merge pull request #5583 from ostezer/docs-ui-add-codeblock-scrolling
Introduce x-axe scrolling to code-blocks.
2014-05-09 10:31:52 +02:00
James Turnbull 12201e2ea3 Merge pull request #5599 from ostezer/docs-rewrite-examples-mongodb
Rewrite and update the MongoDB service article
2014-05-09 10:25:37 +02:00
O.S.Tezer fd56723494 Rewrite and update the MongoDB service article
MongoDB article had some fundemental issues.

 - Outdated Dockerfile
 - Insufficient / unclear instructions
 - Unnecessary comments
 - Failed to explain the role of Docker.io
 - Did not have a complete Dockerfile sample
 - Lacked a "learn more" section / link to Trusted Builds

This update aims to address all these issues with a complete re-write.

It also:

 - Corrects the label under which this article is/was listed on the menu

Docker-DCO-1.1-Signed-off-by: O.S. Tezer <ostezer@gmail.com> (github: ostezer)

 - First run at amending after the initial review process.
 - Make the Dockerfile generic.
 - Revision.
 - Fixes
2014-05-09 09:09:52 +01:00
James Turnbull 2463849ecf Merge pull request #5678 from MasonM/patch-2
Fix link to daemon/execdriver/lxc/lxc_template.go
2014-05-09 09:41:47 +02:00
James Turnbull 80b750be5d Merge pull request #5689 from benatkin/doc-osx-setup-section-fix
make higher level heading specify that instructions are for manual installation
2014-05-09 09:35:24 +02:00
Solomon Hykes d2a41b38c5 Merge pull request #5540 from unclejack/imports_cleanup 2014-05-09 00:30:14 -07:00
unclejack 4c2b9d7324 rename goruntime import to runtime
This renames the goruntime import of the runtime package back to
runtime.

Docker-DCO-1.1-Signed-off-by: Cristian Staretu <cristian.staretu@gmail.com> (github: unclejack)
2014-05-09 10:14:37 +03:00
Solomon Hykes a6b7e94696 Merge pull request #5561 from bmurphy1976/bmurphy1976-table-tests 2014-05-09 00:00:19 -07:00
Solomon Hykes 1736b025ea Merge pull request #5116 from LK4D4/chown_only_on_content_#5110 2014-05-08 23:16:22 -07:00
Benjamin Atkin ca33d2589f make higher level heading specify that instructions are for manual install
Docker-DCO-1.1-Signed-off-by: Benjamin Atkin <ben@benatkin.com> (github: benatkin)
2014-05-08 22:57:04 -07:00
Michael Crosby 170e4d2e19 Cast Input and Output to closer
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-08 12:57:19 -07:00
O.S. Tezer f5309a2a09 Merge pull request #5679 from tdaws/master
Fixed typo in docs - intro - working with docker - working with images. ...
2014-05-08 20:15:54 +01:00
Victor Vieux 8d3c5b7dca Merge pull request #5680 from alexlarsson/dm-remove-spew
devmapper: Remove accidental debug spew
2014-05-08 10:32:11 -07:00
Alexander Larsson bff0c4f3dc devmapper: Remove accidental debug spew
Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-05-08 19:25:47 +02:00
Michael Crosby 7673e3c589 Merge pull request #5673 from tianon/kcore-error
Update restrict.Restrict to both show the error message when failing to mount /dev/null over /proc/kcore, and to ignore "not exists" errors while doing so (for when CONFIG_PROC_KCORE=n in the kernel)
2014-05-08 10:20:19 -07:00
Tony Daws ea049a2ebc Fixed typo in docs - intro - working with docker - working with images. It read that you can use both images used by others and images used by others, in a run-on sentence. 2014-05-08 10:05:00 -07:00
Mason Malone 5d39175c22 Fix link to daemon/execdriver/lxc/lxc_template.go 2014-05-08 12:49:50 -04:00
Michael Crosby 718154b3b6 Merge pull request #5535 from vmarmol/add-maintainers-cgroup
Adding Rohit Jnagal and Victor Marmol to pkg/cgroups maintainers.
2014-05-08 09:48:31 -07:00
Alexander Larsson 3a1703a79f Merge pull request #5404 from alexlarsson/dm-new-metadata
Make devicemapper backend able to support multiple processes
2014-05-08 14:31:17 +02:00
O.S. Tezer e4f4ce83d2 Merge pull request #5659 from ryanaslett/master
Trivial Documentation Change: The Go template is a better pattern for documentation.
2014-05-08 10:59:35 +01:00
Tianon Gravi d60301edb8 Update restrict.Restrict to both show the error message when failing to mount /dev/null over /proc/kcore, and to ignore "not exists" errors while doing so (for when CONFIG_PROC_KCORE=n in the kernel)
Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-08 01:03:45 -06:00
Bryan Murphy fdccfaf72a move Table to a separate file and add additional unit tests
Docker-DCO-1.1-Signed-off-by: Bryan Murphy <bmurphy1976@gmail.com> (github: bmurphy1976)
2014-05-08 02:31:23 +00:00
Victor Vieux 81041b3ba6 Merge pull request #5671 from crosbymichael/merge_release_v0.11.1
Merge release v0.11.1
2014-05-07 18:51:29 -07:00
Michael Crosby 0ed9bba1ee Change version to v0.11.1-dev
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-07 18:48:38 -07:00
Michael Crosby 093aacb740 Merge branch 'release' into merge_release_v0.11.1
Docker-DCO-1.1-Signed-off-by: Michael Crosby <michael@crosbymichael.com> (github: crosbymichael)
2014-05-07 18:47:47 -07:00
Michael Crosby 5f3d6b8e7f Merge pull request #5668 from crosbymichael/bump_v0.11.1
Bump to version v0.11.1
2014-05-07 18:47:46 -07:00
William Henry 7cd64f0344 Merge remote-tracking branch 'upstream/master' 2014-05-07 14:48:28 -06:00
LK4D4 91b7d8ebd3 Change owner only on copied content
Fixes #5110
Docker-DCO-1.1-Signed-off-by: Alexandr Morozov <lk4d4math@gmail.com> (github: LK4D4)
2014-05-07 20:40:22 +04:00
Tianon Gravi 36d913078b Add "Should-Start" cgroupfs-mount and cgroup-lite to sysvinit-debian init script
It's fine to list both here because "Should-Start" is a loose binding (ie, if the listed service exists, it'll be started first, but otherwise, this one will start without it).

Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-05-07 09:46:41 -06:00
Ryan Aslett 7549199cc5 The Go template is a better pattern for documentation. 2014-05-07 00:21:02 -07:00
Vincent Batts 6f32727844 registry: adding vbatts to the MAINTAINERS
Docker-DCO-1.1-Signed-off-by: Vincent Batts <vbatts@redhat.com> (github: vbatts)
2014-05-06 14:31:47 -04:00
O.S.Tezer 02f95a089c Add code for Pardot implementation to docs
Docker-DCO-1.1-Signed-off-by: O.S. Tezer <ostezer@gmail.com> (github: ostezer)
2014-05-06 09:53:19 +01:00
O.S.Tezer dcae8c157b Introduce x-axe scrolling to code-blocks.
This improvement introduces x-axe scrolling to code-blocks by
overriding certain default bootstrap word-wrapping constraints.

This PR closes: https://github.com/dotcloud/docker/issues/5279

Docker-DCO-1.1-Signed-off-by: O.S. Tezer <ostezer@gmail.com> (github: ostezer)
2014-05-04 13:25:03 +01:00
Victor Marmol 91b5fe8502 Adding Rohit Jnagal and Victor Marmol to pkg/cgroups maintainers.
Docker-DCO-1.1-Signed-off-by: Victor Marmol <vmarmol@google.com> (github: vmarmol)
2014-05-01 15:48:16 -07:00
Sven Dowideit b06ad88ca0 Don't add yourself to the AUTHORs file - its generated
Last time I discussed this with @tianon, the AUTHORS was being
regenerated regularly, so we could remove this step.

Docker-DCO-1.1-Signed-off-by: Sven Dowideit <SvenDowideit@fosiki.com> (github: SvenDowideit)
2014-05-01 11:15:07 +10:00
Sven Dowideit 48dbee0824 modernise the MAINTAINER process documentation to line up with what I understand it to be
Docker-DCO-1.1-Signed-off-by: Sven Dowideit <SvenDowideit@fosiki.com> (github: SvenDowideit)
2014-04-29 15:28:04 +10:00
Tianon Gravi 659305085f Update hack/dind to mount cgroups on "/cgroup" instead of "/sys/fs/cgroup" for better compatibility
Fixes #5122

Docker-DCO-1.1-Signed-off-by: Andrew Page <admwiggin@gmail.com> (github: tianon)
2014-04-28 23:17:36 -06:00
Kato Kazuyoshi 21b42dfcfd UTIME_OMIT is only available on Linux
Docker-DCO-1.1-Signed-off-by: Kato Kazuyoshi <kato.kazuyoshi@gmail.com> (github: kzys)
2014-04-29 09:08:33 +09:00
Hobofan 6893689336 reapply Cmd value even if CmdAdd returns early
Docker-DCO-1.1-Signed-off-by: Maximilian Goisser <goisser94@gmail.com> (github: hobofan)
2014-04-29 00:27:32 +02:00
Alexander Larsson 6d631968fa devmapper: Store metadata in one file per device
This allows multiple instances of the backend in different containers
to access devices (although generally only one can modify/create them).

Any old metadata is converted on the first run.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-04-25 15:21:56 +02:00
Alexander Larsson f26203cf54 devmapper: Simplify thin pool device id allocation
Instead of globally keeping track of the free device ids we just
start from 0 each run and handle EEXIST error and try the next one.

This way we don't need any global state for the device ids, which
means we can read device metadata lazily. This is important for
multi-process use of the backend.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-04-25 14:26:27 +02:00
Alexander Larsson 586a511cb5 devmapper: Move error detection to devmapper.go
This moves the EBUSY detection to devmapper.go, and then returns
a real ErrBusy that deviceset uses.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-04-25 14:26:27 +02:00
Alexander Larsson 8116b86e05 devmapper: Maks createSnapDevice a function, not a method
No idea why this was a method. Maybe a cut and paste bug.

Docker-DCO-1.1-Signed-off-by: Alexander Larsson <alexl@redhat.com> (github: alexlarsson)
2014-04-25 14:26:27 +02:00
William Henry 70654160e6 Merge branch 'master' of https://github.com/dotcloud/docker 2014-04-22 22:14:04 -06:00
465 changed files with 18288 additions and 9386 deletions
+29 -9
View File
@@ -6,14 +6,16 @@ Guillaume J. Charmes <guillaume.charmes@docker.com> <charmes.guillaume@gmail.com
<guillaume.charmes@docker.com> <guillaume@dotcloud.com>
<guillaume.charmes@docker.com> <guillaume@docker.com>
<guillaume.charmes@docker.com> <guillaume.charmes@dotcloud.com>
<guillaume.charmes@docker.com> <guillaume@charmes.net>
<kencochrane@gmail.com> <KenCochrane@gmail.com>
<sridharr@activestate.com> <github@srid.name>
Thatcher Peskens <thatcher@dotcloud.com> dhrp <thatcher@dotcloud.com>
Thatcher Peskens <thatcher@dotcloud.com> dhrp <thatcher@gmx.net>
Thatcher Peskens <thatcher@docker.com>
Thatcher Peskens <thatcher@docker.com> <thatcher@dotcloud.com>
Thatcher Peskens <thatcher@docker.com> dhrp <thatcher@gmx.net>
Jérôme Petazzoni <jerome.petazzoni@dotcloud.com> jpetazzo <jerome.petazzoni@dotcloud.com>
Jérôme Petazzoni <jerome.petazzoni@dotcloud.com> <jp@enix.org>
Joffrey F <joffrey@dotcloud.com>
<joffrey@dotcloud.com> <f.joffrey@gmail.com>
Joffrey F <joffrey@docker.com>
Joffrey F <joffrey@docker.com> <joffrey@dotcloud.com>
Joffrey F <joffrey@docker.com> <f.joffrey@gmail.com>
Tim Terhorst <mynamewastaken+git@gmail.com>
Andy Smith <github@anarkystic.com>
<kalessin@kalessin.fr> <louis@dotcloud.com>
@@ -23,7 +25,6 @@ Andy Smith <github@anarkystic.com>
<victor.vieux@docker.com> <victor@docker.com>
<victor.vieux@docker.com> <vieux@docker.com>
<dominik@honnef.co> <dominikh@fork-bomb.org>
Thatcher Peskens <thatcher@dotcloud.com>
<ehanchrow@ine.com> <eric.hanchrow@gmail.com>
Walter Stanish <walter@pratyeka.org>
<daniel@gasienica.ch> <dgasienica@zynga.com>
@@ -54,7 +55,26 @@ Jean-Baptiste Dalido <jeanbaptiste@appgratis.com>
<gurjeet@singh.im> <singh.gurjeet@gmail.com>
<shawn@churchofgit.com> <shawnlandden@gmail.com>
<sjoerd-github@linuxonly.nl> <sjoerd@byte.nl>
<solomon@dotcloud.com> <solomon.hykes@dotcloud.com>
<SvenDowideit@home.org.au> <SvenDowideit@fosiki.com>
Sven Dowideit <SvenDowideit@home.org.au> ¨Sven <¨SvenDowideit@home.org.au¨>
<solomon@docker.com> <solomon.hykes@dotcloud.com>
<solomon@docker.com> <solomon@dotcloud.com>
Sven Dowideit <SvenDowideit@home.org.au>
Sven Dowideit <SvenDowideit@home.org.au> <SvenDowideit@fosiki.com>
Sven Dowideit <SvenDowideit@home.org.au> <SvenDowideit@docker.com>
Sven Dowideit <SvenDowideit@home.org.au> <¨SvenDowideit@home.org.au¨>
unclejack <unclejacksons@gmail.com> <unclejack@users.noreply.github.com>
<alexl@redhat.com> <alexander.larsson@gmail.com>
Alexandr Morozov <lk4d4math@gmail.com>
<git.nivoc@neverbox.com> <kuehnle@online.de>
O.S. Tezer <ostezer@gmail.com>
<ostezer@gmail.com> <ostezer@users.noreply.github.com>
Roberto G. Hashioka <roberto.hashioka@docker.com> <roberto_hashioka@hotmail.com>
<justin.p.simonelis@gmail.com> <justin.simonelis@PTS-JSIMON2.toronto.exclamation.com>
<taim@bosboot.org> <maztaim@users.noreply.github.com>
<viktor.vojnovski@amadeus.com> <vojnovski@gmail.com>
<vbatts@redhat.com> <vbatts@hashbangbash.com>
<altsysrq@gmail.com> <iamironbob@gmail.com>
Sridhar Ratnakumar <sridharr@activestate.com>
Sridhar Ratnakumar <sridharr@activestate.com> <github@srid.name>
Liang-Chi Hsieh <viirya@gmail.com>
Aleksa Sarai <cyphar@cyphar.com>
Will Weaver <monkey@buildingbananas.com>
+120 -13
View File
@@ -1,44 +1,62 @@
# This file lists all individuals having contributed content to the repository.
# If you're submitting a patch, please add your name here in alphabetical order as part of the patch.
#
# For a list of active project maintainers, see the MAINTAINERS file.
#
# For how it is generated, see `.mailmap`.
Aanand Prasad <aanand.prasad@gmail.com>
Aaron Feng <aaron.feng@gmail.com>
Aaron Huslage <huslage@gmail.com>
Abel Muiño <amuino@gmail.com>
Adam Miller <admiller@redhat.com>
Adam Singer <financeCoding@gmail.com>
Aditya <aditya@netroy.in>
Adrian Mouat <adrian.mouat@gmail.com>
alambike <alambike@gmail.com>
Aleksa Sarai <cyphar@cyphar.com>
Alexander Larsson <alexl@redhat.com>
Alexandr Morozov <lk4d4math@gmail.com>
Alexey Kotlyarov <alexey@infoxchange.net.au>
Alexey Shamrin <shamrin@gmail.com>
Alex Gaynor <alex.gaynor@gmail.com>
Alexis THOMAS <fr.alexisthomas@gmail.com>
almoehi <almoehi@users.noreply.github.com>
Al Tobey <al@ooyala.com>
amangoel <amangoel@gmail.com>
Andrea Luzzardi <aluzzardi@gmail.com>
Andreas Savvides <andreas@editd.com>
Andreas Tiefenthaler <at@an-ti.eu>
Andrea Turli <andrea.turli@gmail.com>
Andrew Duckworth <grillopress@gmail.com>
Andrew Macgregor <andrew.macgregor@agworld.com.au>
Andrew Munsell <andrew@wizardapps.net>
Andrews Medina <andrewsmedina@gmail.com>
Andrew Williams <williams.andrew@gmail.com>
Andy Chambers <anchambers@paypal.com>
andy diller <dillera@gmail.com>
Andy Goldstein <agoldste@redhat.com>
Andy Kipp <andy@rstudio.com>
Andy Rothfusz <github@metaliveblog.com>
Andy Smith <github@anarkystic.com>
Anthony Bishopric <git@anthonybishopric.com>
Anton Nikitin <anton.k.nikitin@gmail.com>
Antony Messerli <amesserl@rackspace.com>
apocas <petermdias@gmail.com>
Arnaud Porterie <icecrime@gmail.com>
Asbjørn Enge <asbjorn@hanafjedle.net>
Barnaby Gray <barnaby@pickle.me.uk>
Barry Allard <barry.allard@gmail.com>
Bartłomiej Piotrowski <b@bpiotrowski.pl>
Benjamin Atkin <ben@benatkin.com>
Benoit Chesneau <bchesneau@gmail.com>
Ben Sargent <ben@brokendigits.com>
Ben Toews <mastahyeti@gmail.com>
Ben Wiklund <ben@daisyowl.com>
Bernerd Schaefer <bj.schaefer@gmail.com>
Bhiraj Butala <abhiraj.butala@gmail.com>
bin liu <liubin0329@users.noreply.github.com>
Bouke Haarsma <bouke@webatoom.nl>
Brandon Liu <bdon@bdon.org>
Brandon Philips <brandon@ifup.org>
Brian Dorsey <brian@dorseys.org>
Brian Flad <bflad417@gmail.com>
Brian Goff <cpuguy83@gmail.com>
Brian McCallister <brianm@skife.org>
Brian Olsen <brian@maven-group.org>
@@ -46,11 +64,15 @@ Brian Shumate <brian@couchbase.com>
Briehan Lombaard <briehan.lombaard@gmail.com>
Bruno Bigras <bigras.bruno@gmail.com>
Bryan Matsuo <bryan.matsuo@gmail.com>
Bryan Murphy <bmurphy1976@gmail.com>
Caleb Spare <cespare@gmail.com>
Calen Pennington <cale@edx.org>
Cameron Boehmer <cameron.boehmer@gmail.com>
Carl X. Su <bcbcarl@gmail.com>
Charles Hooper <charles.hooper@dotcloud.com>
Charles Lindsay <chaz@chazomatic.us>
Charles Merriam <charles.merriam@gmail.com>
Charlie Lewis <charliel@lab41.org>
Chia-liang Kao <clkao@clkao.org>
Chris St. Pierre <chris.a.st.pierre@gmail.com>
Christopher Currie <codemonkey+github@gmail.com>
@@ -61,6 +83,7 @@ Colin Dunklau <colin.dunklau@gmail.com>
Colin Rice <colin@daedrum.net>
Cory Forsyth <cory.forsyth@gmail.com>
cressie176 <github@stephen-cresswell.net>
Dafydd Crosby <dtcrsby@gmail.com>
Dan Buch <d.buch@modcloth.com>
Dan Hirsch <thequux@upstandinghackers.com>
Daniel Exner <dex@dragonslave.de>
@@ -72,30 +95,45 @@ Daniel Nordberg <dnordberg@gmail.com>
Daniel Robinson <gottagetmac@gmail.com>
Daniel Von Fange <daniel@leancoder.com>
Daniel YC Lin <dlin.tw@gmail.com>
Dan Keder <dan.keder@gmail.com>
Dan McPherson <dmcphers@redhat.com>
Danny Berger <dpb587@gmail.com>
Danny Yates <danny@codeaholics.org>
Dan Stine <sw@stinemail.com>
Dan Walsh <dwalsh@redhat.com>
Dan Williams <me@deedubs.com>
Darren Coxall <darren@darrencoxall.com>
Darren Shepherd <darren.s.shepherd@gmail.com>
David Anderson <dave@natulte.net>
David Calavera <david.calavera@gmail.com>
David Gageot <david@gageot.net>
David Mcanulty <github@hellspark.com>
David Röthlisberger <david@rothlis.net>
David Sissitka <me@dsissitka.com>
Deni Bertovic <deni@kset.org>
Dinesh Subhraveti <dineshs@altiscale.com>
Djibril Koné <kone.djibril@gmail.com>
dkumor <daniel@dkumor.com>
Dmitry Demeshchuk <demeshchuk@gmail.com>
Dolph Mathews <dolph.mathews@gmail.com>
Dominik Honnef <dominik@honnef.co>
Don Spaulding <donspauldingii@gmail.com>
Dražen Lučanin <kermit666@gmail.com>
Dr Nic Williams <drnicwilliams@gmail.com>
Dustin Sallings <dustin@spy.net>
Edmund Wagner <edmund-wagner@web.de>
Eiichi Tsukata <devel@etsukata.com>
Eivind Uggedal <eivind@uggedal.com>
Elias Probst <mail@eliasprobst.eu>
Emil Hernvall <emil@quench.at>
Emily Rose <emily@contactvibe.com>
Eric Hanchrow <ehanchrow@ine.com>
Eric Lee <thenorthsecedes@gmail.com>
Eric Myhre <hash@exultant.us>
Erik Hollensbe <erik+github@hollensbe.org>
Erno Hopearuoho <erno.hopearuoho@gmail.com>
eugenkrizo <eugen.krizo@gmail.com>
Evan Hazlett <ejhazlett@gmail.com>
Evan Krall <krall@yelp.com>
Evan Phoenix <evan@fallingsnow.net>
Evan Wies <evan@neomantra.net>
@@ -106,6 +144,7 @@ Fabio Rehm <fgrehm@gmail.com>
Fabrizio Regini <freegenie@gmail.com>
Faiz Khan <faizkhan00@gmail.com>
Fareed Dudhia <fareeddudhia@googlemail.com>
Felix Rabe <felix@rabe.io>
Fernando <fermayo@gmail.com>
Flavio Castelli <fcastelli@suse.com>
Francisco Souza <f@souza.cc>
@@ -117,8 +156,11 @@ Gabe Rosenhouse <gabe@missionst.com>
Gabriel Monroy <gabriel@opdemand.com>
Galen Sampson <galen.sampson@gmail.com>
Gareth Rushgrove <gareth@morethanseven.net>
Geoffrey Bachelet <grosfrais@gmail.com>
Gereon Frey <gereon.frey@dynport.de>
German DZ <germ@ndz.com.ar>
Gert van Valkenhoef <g.h.m.van.valkenhoef@rug.nl>
Goffert van Gool <goffert@phusion.nl>
Graydon Hoare <graydon@pobox.com>
Greg Thornton <xdissent@me.com>
grunny <mwgrunny@gmail.com>
@@ -127,28 +169,40 @@ Gurjeet Singh <gurjeet@singh.im>
Guruprasad <lgp171188@gmail.com>
Harley Laue <losinggeneration@gmail.com>
Hector Castro <hectcastro@gmail.com>
Hobofan <goisser94@gmail.com>
Hunter Blanks <hunter@twilio.com>
Ian Truslove <ian.truslove@gmail.com>
ILYA Khlopotov <ilya.khlopotov@gmail.com>
inglesp <peter.inglesby@gmail.com>
Isaac Dupree <antispam@idupree.com>
Isabel Jimenez <contact.isabeljimenez@gmail.com>
Isao Jonas <isao.jonas@gmail.com>
Jack Danger Canty <jackdanger@squareup.com>
jakedt <jake@devtable.com>
Jake Moshenko <jake@devtable.com>
James Allen <jamesallen0108@gmail.com>
James Carr <james.r.carr@gmail.com>
James DeFelice <james.defelice@ishisystems.com>
James Harrison Fisher <jameshfisher@gmail.com>
James Mills <prologic@shortcircuit.net.au>
James Turnbull <james@lovedthanlost.net>
jaseg <jaseg@jaseg.net>
Jason McVetta <jason.mcvetta@gmail.com>
Jason Plum <jplum@devonit.com>
Jean-Baptiste Barth <jeanbaptiste.barth@gmail.com>
Jean-Baptiste Dalido <jeanbaptiste@appgratis.com>
Jeff Lindsay <progrium@gmail.com>
Jeremy Grosser <jeremy@synack.me>
Jérôme Petazzoni <jerome.petazzoni@dotcloud.com>
Jesse Dubay <jesse@thefortytwo.net>
Jilles Oldenbeuving <ojilles@gmail.com>
Jim Alateras <jima@comware.com.au>
Jimmy Cuadra <jimmy@jimmycuadra.com>
Joe Beda <joe.github@bedafamily.com>
Joel Handwell <joelhandwell@gmail.com>
Joe Shaw <joe@joeshaw.org>
Joe Van Dyk <joe@tanga.com>
Joffrey F <joffrey@dotcloud.com>
Joffrey F <joffrey@docker.com>
Johan Euphrosine <proppy@google.com>
Johannes 'fish' Ziemke <github@freigeist.org>
Johan Rydberg <johan.rydberg@gmail.com>
@@ -157,7 +211,9 @@ John Feminella <jxf@jxf.me>
John Gardiner Myers <jgmyers@proofpoint.com>
John Warwick <jwarwick@gmail.com>
Jonas Pfenniger <jonas@pfenniger.name>
Jonathan McCrohan <jmccrohan@gmail.com>
Jonathan Mueller <j.mueller@apoveda.ch>
Jonathan Pares <jonathanpa@users.noreply.github.com>
Jonathan Rudenberg <jonathan@titanous.com>
Jon Wedaman <jweede@gmail.com>
Joost Cassee <joost@cassee.net>
@@ -172,13 +228,17 @@ Julien Barbier <write0@gmail.com>
Julien Dubois <julien.dubois@gmail.com>
Justin Force <justin.force@gmail.com>
Justin Plock <jplock@users.noreply.github.com>
Justin Simonelis <justin.p.simonelis@gmail.com>
Karan Lyons <karan@karanlyons.com>
Karl Grzeszczak <karlgrz@gmail.com>
Kato Kazuyoshi <kato.kazuyoshi@gmail.com>
Kawsar Saiyeed <kawsar.saiyeed@projiris.com>
Keli Hu <dev@keli.hu>
Ken Cochrane <kencochrane@gmail.com>
Ken ICHIKAWA <ichikawa.ken@jp.fujitsu.com>
Kevin Clark <kevin.clark@gmail.com>
Kevin J. Lynagh <kevin@keminglabs.com>
Kevin Menard <kevin@nirvdrum.com>
Kevin Wallace <kevin@pentabarf.net>
Keyvan Fatehi <keyvanfatehi@gmail.com>
kim0 <email.ahmedkamal@googlemail.com>
@@ -187,14 +247,20 @@ Kimbro Staken <kstaken@kstaken.com>
Kiran Gangadharan <kiran.daredevil@gmail.com>
Konstantin Pelykh <kpelykh@zettaset.com>
Kyle Conroy <kyle.j.conroy@gmail.com>
lalyos <lalyos@yahoo.com>
Lance Chen <cyen0312@gmail.com>
Lars R. Damerow <lars@pixar.com>
Laurie Voss <github@seldo.com>
Lewis Peckover <lew+github@lew.io>
Liang-Chi Hsieh <viirya@gmail.com>
Lokesh Mandvekar <lsm5@redhat.com>
Louis Opter <kalessin@kalessin.fr>
lukaspustina <lukas.pustina@centerdevice.com>
lukemarsden <luke@digital-crocus.com>
Mahesh Tiyyagura <tmahesh@gmail.com>
Manuel Meurer <manuel@krautcomputing.com>
Manuel Woelker <github@manuel.woelker.org>
Marc Abramowitz <marc@marc-abramowitz.com>
Marc Kuo <kuomarc2@gmail.com>
Marco Hennings <marco.hennings@freiheit.com>
Marcus Farkas <toothlessgear@finitebox.com>
@@ -206,23 +272,32 @@ Marko Mikulicic <mmikulicic@gmail.com>
Markus Fix <lispmeister@gmail.com>
Martijn van Oosterhout <kleptog@svana.org>
Martin Redmond <martin@tinychat.com>
Mason Malone <mason.malone@gmail.com>
Mateusz Sulima <sulima.mateusz@gmail.com>
Mathieu Le Marec - Pasquet <kiorky@cryptelium.net>
Matt Apperson <me@mattapperson.com>
Matt Bachmann <bachmann.matt@gmail.com>
Matt Haggard <haggardii@gmail.com>
Matthew Mueller <mattmuelle@gmail.com>
Matthias Klumpp <matthias@tenstral.net>
Matthias Kühnle <git.nivoc@neverbox.com>
mattymo <raytrac3r@gmail.com>
Maxime Petazzoni <max@signalfuse.com>
Maxim Treskin <zerthurd@gmail.com>
Max Shytikov <mshytikov@gmail.com>
meejah <meejah@meejah.ca>
Michael Brown <michael@netdirect.ca>
Michael Crosby <michael@crosbymichael.com>
Michael Gorsuch <gorsuch@github.com>
Michael Neale <michael.neale@gmail.com>
Michael Stapelberg <michael+gh@stapelberg.de>
Miguel Angel Fernández <elmendalerenda@gmail.com>
Mike Gaffney <mike@uberu.com>
Mike MacCana <mike.maccana@gmail.com>
Mike Naberezny <mike@naberezny.com>
Mikhail Sobolev <mss@mawhrin.net>
Mohit Soni <mosoni@ebay.com>
Morgante Pell <morgante.pell@morgante.net>
Morten Siebuhr <sbhr@sbhr.dk>
Nan Monnand Deng <monnand@gmail.com>
Nate Jones <nate@endot.org>
@@ -234,22 +309,26 @@ Nick Stenning <nick.stenning@digital.cabinet-office.gov.uk>
Nick Stinemates <nick@stinemates.org>
Nicolas Dudebout <nicolas.dudebout@gatech.edu>
Nicolas Kaiser <nikai@nikai.net>
noducks <onemannoducks@gmail.com>
Nolan Darilek <nolan@thewordnerd.info>
odk- <github@odkurzacz.org>
Oguz Bilgic <fisyonet@gmail.com>
Ole Reifschneider <mail@ole-reifschneider.de>
O.S.Tezer <ostezer@gmail.com>
O.S. Tezer <ostezer@gmail.com>
pandrew <letters@paulnotcom.se>
Pascal Borreli <pascal@borreli.com>
pattichen <craftsbear@gmail.com>
Paul Annesley <paul@annesley.cc>
Paul Bowsher <pbowsher@globalpersonals.co.uk>
Paul Hammond <paul@paulhammond.org>
Paul Jimenez <pj@place.org>
Paul Lietar <paul@lietar.net>
Paul Morie <pmorie@gmail.com>
Paul Nasrat <pnasrat@gmail.com>
Paul <paul9869@gmail.com>
Peter Braden <peterbraden@peterbraden.co.uk>
Peter Waller <peter@scraperwiki.com>
Phillip Alexander <git@phillipalexander.io>
Phil Spitler <pspitler@gmail.com>
Piergiuliano Bossi <pgbossi@gmail.com>
Pierre-Alain RIVIERE <pariviere@ippon.fr>
@@ -257,6 +336,8 @@ Piotr Bogdan <ppbogdan@gmail.com>
pysqz <randomq@126.com>
Quentin Brossard <qbrossard@gmail.com>
Rafal Jeczalik <rjeczalik@gmail.com>
Rajat Pandit <rp@rajatpandit.com>
Ralph Bean <rbean@redhat.com>
Ramkumar Ramachandra <artagnon@gmail.com>
Ramon van Alteren <ramon@vanalteren.nl>
Renato Riccieri Santos Zannon <renato.riccieri@gmail.com>
@@ -266,54 +347,71 @@ Richo Healey <richo@psych0tik.net>
Rick Bradley <rick@users.noreply.github.com>
Robert Obryk <robryk@gmail.com>
Roberto G. Hashioka <roberto.hashioka@docker.com>
Roberto Hashioka <roberto_hashioka@hotmail.com>
robpc <rpcann@gmail.com>
Rodrigo Vaz <rodrigo.vaz@gmail.com>
Roel Van Nyen <roel.vannyen@gmail.com>
Roger Peppe <rogpeppe@gmail.com>
Rohit Jnagal <jnagal@google.com>
Roland Moriz <rmoriz@users.noreply.github.com>
Rovanion Luckey <rovanion.luckey@gmail.com>
Ryan Aslett <github@mixologic.com>
Ryan Fowler <rwfowler@gmail.com>
Ryan O'Donnell <odonnellryanc@gmail.com>
Ryan Seto <ryanseto@yak.net>
Ryan Thomas <rthomas@atlassian.com>
Sam Alba <sam.alba@gmail.com>
Sam J Sharpe <sam.sharpe@digital.cabinet-office.gov.uk>
Sam Rijs <srijs@airpost.net>
Samuel Andaya <samuel@andaya.net>
Scott Bessler <scottbessler@gmail.com>
Scott Collier <emailscottcollier@gmail.com>
Sean Cronin <seancron@gmail.com>
Sean P. Kane <skane@newrelic.com>
Sébastien Stormacq <sebsto@users.noreply.github.com>
Shawn Landden <shawn@churchofgit.com>
Shawn Siefkas <shawn.siefkas@meredith.com>
Shih-Yuan Lee <fourdollars@gmail.com>
shin- <joffrey@docker.com>
Silas Sewell <silas@sewell.org>
Simon Taranto <simon.taranto@gmail.com>
Sindhu S <sindhus@live.in>
Sjoerd Langkemper <sjoerd-github@linuxonly.nl>
Solomon Hykes <solomon@dotcloud.com>
Solomon Hykes <solomon@docker.com>
Song Gao <song@gao.io>
Soulou <leo@unbekandt.eu>
Sridatta Thatipamala <sthatipamala@gmail.com>
Sridhar Ratnakumar <sridharr@activestate.com>
Steeve Morin <steeve.morin@gmail.com>
Stefan Praszalowicz <stefan@greplin.com>
Steven Burgess <steven.a.burgess@hotmail.com>
sudosurootdev <sudosurootdev@gmail.com>
Sven Dowideit <svendowideit@home.org.au>
Sven Dowideit <SvenDowideit@home.org.au>
Sylvain Bellemare <sylvain.bellemare@ezeep.com>
tang0th <tang0th@gmx.com>
Tatsuki Sugiura <sugi@nemui.org>
Tehmasp Chaudhri <tehmasp@gmail.com>
Thatcher Peskens <thatcher@dotcloud.com>
Thatcher Peskens <thatcher@docker.com>
Thermionix <bond711@gmail.com>
Thijs Terlouw <thijsterlouw@gmail.com>
Thomas Bikeev <thomas.bikeev@mac.com>
Thomas Frössman <thomasf@jossystem.se>
Thomas Hansen <thomas.hansen@gmail.com>
Thomas LEVEIL <thomasleveil@gmail.com>
Thomas Schroeter <thomas@cliqz.com>
Tianon Gravi <admwiggin@gmail.com>
Tim Bosse <maztaim@users.noreply.github.com>
Tibor Vass <teabee89@gmail.com>
Tim Bosse <taim@bosboot.org>
Timothy Hobbs <timothyhobbs@seznam.cz>
Tim Ruffles <oi@truffles.me.uk>
Tim Terhorst <mynamewastaken+git@gmail.com>
tjmehta <tj@init.me>
Tobias Bieniek <Tobias.Bieniek@gmx.de>
Tobias Schmidt <ts@soundcloud.com>
Tobias Schwab <tobias.schwab@dynport.de>
Todd Lunter <tlunter@gmail.com>
Tom Fotherby <tom+github@peopleperhour.com>
Tom Hulihan <hulihan.tom159@gmail.com>
Tommaso Visconti <tommaso.visconti@gmail.com>
Tony Daws <tony@daws.ca>
Travis Cline <travis.cline@gmail.com>
Tyler Brock <tyler.brock@gmail.com>
Tzu-Jung Lee <roylee17@gmail.com>
@@ -322,26 +420,35 @@ unclejack <unclejacksons@gmail.com>
vgeta <gopikannan.venugopalsamy@gmail.com>
Victor Coisne <victor.coisne@dotcloud.com>
Victor Lyuboslavsky <victor@victoreda.com>
Victor Marmol <vmarmol@google.com>
Victor Vieux <victor.vieux@docker.com>
Viktor Vojnovski <viktor.vojnovski@amadeus.com>
Vincent Batts <vbatts@redhat.com>
Vincent Bernat <bernat@luffy.cx>
Vincent Mayers <vincent.mayers@inbloom.org>
Vincent Woo <me@vincentwoo.com>
Vinod Kulkarni <vinod.kulkarni@gmail.com>
Vishnu Kannan <vishnuk@google.com>
Vitor Monteiro <vmrmonteiro@gmail.com>
Vivek Agarwal <me@vivek.im>
Vladimir Bulyga <xx@ccxx.cc>
Vladimir Kirillov <proger@wilab.org.ua>
Vladimir Rutsky <iamironbob@gmail.com>
Vladimir Rutsky <altsysrq@gmail.com>
Walter Leibbrandt <github@wrl.co.za>
Walter Stanish <walter@pratyeka.org>
WarheadsSE <max@warheads.net>
Wes Morgan <cap10morgan@gmail.com>
Will Dietz <w@wdtz.org>
William Delanoue <william.delanoue@gmail.com>
William Henry <whenry@redhat.com>
Will Rouesnel <w.rouesnel@gmail.com>
Will Weaver <monkey@buildingbananas.com>
Xiuming Chen <cc@cxm.cc>
Yang Bai <hamo.by@gmail.com>
Yasunori Mahata <nori@mahata.net>
Yurii Rashkovskii <yrashk@gmail.com>
Zain Memon <zain@inzain.net>
Zaiste! <oh@zaiste.net>
Zilin Du <zilin.du@gmail.com>
zimbatm <zimbatm@zimbatm.com>
zqh <zqhxuyuan@gmail.com>
+12
View File
@@ -1,5 +1,17 @@
# Changelog
## 0.12.0 (2014-06-05)
#### Notable features since 0.11.0
* New `COPY` Dockerfile instruction to allow copying a local file from the context into the container without ever extracting if the file is a tar file
* Inherit file permissions from the host on `ADD`
* New `pause` and `unpause` commands to allow pausing and unpausing of containers using cgroup freezer
* The `images` command has a `-f`/`--filter` option to filter the list of images
* Add `--force-rm` to clean up after a failed build
* Standardize JSON keys in Remote API to CamelCase
* Pull from a docker run now assumes `latest` tag if not specified
* Enhance security on Linux capabilities and device nodes
## 0.11.1 (2014-05-07)
#### Registry
+9 -13
View File
@@ -77,13 +77,8 @@ well as a clean documentation build. See ``docs/README.md`` for more
information on building the docs and how docs get released.
Write clean code. Universally formatted code promotes ease of writing, reading,
and maintenance. Always run `go fmt` before committing your changes. Most
editors have plugins that do this automatically, and there's also a git
pre-commit hook:
```
curl -o .git/hooks/pre-commit https://raw.githubusercontent.com/edsrzf/gofmt-git-hook/master/fmt-check && chmod +x .git/hooks/pre-commit
```
and maintenance. Always run `gofmt -s -w file.go` on each changed file before
committing your changes. Most editors have plugins that do this automatically.
Pull requests descriptions should be as clear as possible and include a
reference to all the issues that they address.
@@ -108,10 +103,8 @@ same commit so that a revert would remove all traces of the feature or fix.
Commits that fix or close an issue should include a reference like `Closes #XXX`
or `Fixes #XXX`, which will automatically close the issue when merged.
Add your name to the AUTHORS file, but make sure the list is sorted and your
name and email address match your git configuration. The AUTHORS file is
regenerated occasionally from the git commit history, so a mismatch may result
in your changes being overwritten.
Please do not add yourself to the AUTHORS file, as it is regenerated
regularly from the Git history.
### Merge approval
@@ -182,7 +175,7 @@ One way to automate this, is customise your get ``commit.template`` by adding
a ``prepare-commit-msg`` hook to your docker checkout:
```
curl -o .git/hooks/prepare-commit-msg https://raw.github.com/dotcloud/docker/master/contrib/prepare-commit-msg.hook && chmod +x .git/hooks/prepare-commit-msg
curl -o .git/hooks/prepare-commit-msg https://raw.githubusercontent.com/dotcloud/docker/master/contrib/prepare-commit-msg.hook && chmod +x .git/hooks/prepare-commit-msg
```
* Note: the above script expects to find your GitHub user name in ``git config --get github.user``
@@ -192,7 +185,10 @@ curl -o .git/hooks/prepare-commit-msg https://raw.github.com/dotcloud/docker/mas
There are several exceptions to the signing requirement. Currently these are:
* Your patch fixes spelling or grammar errors.
* Your patch is a single line change to documentation.
* Your patch is a single line change to documentation contained in the
`docs` directory.
* Your patch fixes Markdown formatting or syntax errors in the
documentation contained in the `docs` directory.
If you have any questions, please refer to the FAQ in the [docs](http://docs.docker.io)
+7 -6
View File
@@ -24,7 +24,7 @@
#
docker-version 0.6.1
FROM ubuntu:13.10
FROM ubuntu:14.04
MAINTAINER Tianon Gravi <admwiggin@gmail.com> (@tianon)
# Packaged dependencies
@@ -41,6 +41,7 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -yq \
libapparmor-dev \
libcap-dev \
libsqlite3-dev \
lxc=1.0* \
mercurial \
pandoc \
reprepro \
@@ -49,10 +50,6 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -yq \
s3cmd=1.1.0* \
--no-install-recommends
# Get and compile LXC 0.8 (since it is the most stable)
RUN git clone --no-checkout https://github.com/lxc/lxc.git /usr/local/lxc && cd /usr/local/lxc && git checkout -q lxc-0.8.0
RUN cd /usr/local/lxc && ./autogen.sh && ./configure --disable-docs && make && make install
# Get lvm2 source for compiling statically
RUN git clone --no-checkout https://git.fedorahosted.org/git/lvm2.git /usr/local/lvm2 && cd /usr/local/lvm2 && git checkout -q v2_02_103
# see https://git.fedorahosted.org/cgit/lvm2.git/refs/tags for release tags
@@ -84,7 +81,7 @@ RUN go get code.google.com/p/go.tools/cmd/cover
RUN gem install --no-rdoc --no-ri fpm --version 1.0.2
# Get the "busybox" image source so we can build locally instead of pulling
RUN git clone https://github.com/jpetazzo/docker-busybox.git /docker-busybox
RUN git clone -b buildroot-2014.02 https://github.com/jpetazzo/docker-busybox.git /docker-busybox
# Setup s3cmd config
RUN /bin/echo -e '[default]\naccess_key=$AWS_ACCESS_KEY\nsecret_key=$AWS_SECRET_KEY' > /.s3cfg
@@ -92,6 +89,10 @@ RUN /bin/echo -e '[default]\naccess_key=$AWS_ACCESS_KEY\nsecret_key=$AWS_SECRET_
# Set user.email so crosbymichael's in-container merge commits go smoothly
RUN git config --global user.email 'docker-dummy@example.com'
# Add an unprivileged user to be used for tests which need it
RUN groupadd -r docker
RUN useradd --create-home --gid docker unprivilegeduser
VOLUME /var/lib/docker
WORKDIR /go/src/github.com/dotcloud/docker
ENV DOCKER_BUILDTAGS apparmor selinux
+2
View File
@@ -2,6 +2,8 @@ Solomon Hykes <solomon@docker.com> (@shykes)
Guillaume J. Charmes <guillaume@docker.com> (@creack)
Victor Vieux <vieux@docker.com> (@vieux)
Michael Crosby <michael@crosbymichael.com> (@crosbymichael)
.mailmap: Tianon Gravi <admwiggin@gmail.com> (@tianon)
.travis.yml: Tianon Gravi <admwiggin@gmail.com> (@tianon)
AUTHORS: Tianon Gravi <admwiggin@gmail.com> (@tianon)
Dockerfile: Tianon Gravi <admwiggin@gmail.com> (@tianon)
Makefile: Tianon Gravi <admwiggin@gmail.com> (@tianon)
+1 -1
View File
@@ -35,7 +35,7 @@ docs-release: docs-build
$(DOCKER_RUN_DOCS) "$(DOCKER_DOCS_IMAGE)" ./release.sh
test: build
$(DOCKER_RUN_DOCKER) hack/make.sh binary test-unit test-integration test-integration-cli
$(DOCKER_RUN_DOCKER) hack/make.sh binary cross test-unit test-integration test-integration-cli
test-unit: build
$(DOCKER_RUN_DOCKER) hack/make.sh test-unit
+6
View File
@@ -190,3 +190,9 @@ It is your responsibility to ensure that your use and/or transfer does not
violate applicable laws.
For more information, please see http://www.bis.doc.gov
Licensing
=========
Docker is licensed under the Apache License, Version 2.0. See LICENSE for full license text.
+1 -1
View File
@@ -1 +1 @@
0.11.1
0.12.0
+5
View File
@@ -0,0 +1,5 @@
This directory contains code pertaining to the Docker API:
- Used by the docker client when comunicating with the docker deamon
- Used by third party tools wishing to interface with the docker deamon
+4 -1
View File
@@ -23,6 +23,9 @@ var funcMap = template.FuncMap{
}
func (cli *DockerCli) getMethod(name string) (func(...string) error, bool) {
if len(name) == 0 {
return nil, false
}
methodName := "Cmd" + strings.ToUpper(name[:1]) + strings.ToLower(name[1:])
method := reflect.ValueOf(cli).MethodByName(methodName)
if !method.IsValid() {
@@ -73,7 +76,7 @@ func NewDockerCli(in io.ReadCloser, out, err io.Writer, proto, addr string, tlsC
}
if in != nil {
if file, ok := in.(*os.File); ok {
if file, ok := out.(*os.File); ok {
terminalFd = file.Fd()
isTerminal = term.IsTerminal(terminalFd)
}
+191 -82
View File
@@ -13,7 +13,7 @@ import (
"os"
"os/exec"
"path"
goruntime "runtime"
"runtime"
"strconv"
"strings"
"syscall"
@@ -26,11 +26,14 @@ import (
"github.com/dotcloud/docker/dockerversion"
"github.com/dotcloud/docker/engine"
"github.com/dotcloud/docker/nat"
"github.com/dotcloud/docker/opts"
"github.com/dotcloud/docker/pkg/signal"
"github.com/dotcloud/docker/pkg/term"
"github.com/dotcloud/docker/pkg/units"
"github.com/dotcloud/docker/registry"
"github.com/dotcloud/docker/runconfig"
"github.com/dotcloud/docker/utils"
"github.com/dotcloud/docker/utils/filters"
)
func (cli *DockerCli) CmdHelp(args ...string) error {
@@ -46,7 +49,7 @@ func (cli *DockerCli) CmdHelp(args ...string) error {
help := fmt.Sprintf("Usage: docker [OPTIONS] COMMAND [arg...]\n -H=[unix://%s]: tcp://host:port to bind/connect to or unix://path/to/socket to use\n\nA self-sufficient runtime for linux containers.\n\nCommands:\n", api.DEFAULTUNIXSOCKET)
for _, command := range [][]string{
{"attach", "Attach to a running container"},
{"build", "Build a container from a Dockerfile"},
{"build", "Build an image from a Dockerfile"},
{"commit", "Create a new image from a container's changes"},
{"cp", "Copy files/folders from the containers filesystem to the host path"},
{"diff", "Inspect changes on a container's filesystem"},
@@ -62,6 +65,7 @@ func (cli *DockerCli) CmdHelp(args ...string) error {
{"login", "Register or Login to the docker registry server"},
{"logs", "Fetch the logs of a container"},
{"port", "Lookup the public-facing port which is NAT-ed to PRIVATE_PORT"},
{"pause", "Pause all processes within a container"},
{"ps", "List containers"},
{"pull", "Pull an image or a repository from the docker registry server"},
{"push", "Push an image or a repository to the docker registry server"},
@@ -75,6 +79,7 @@ func (cli *DockerCli) CmdHelp(args ...string) error {
{"stop", "Stop a running container"},
{"tag", "Tag an image into a repository"},
{"top", "Lookup the running processes of a container"},
{"unpause", "Unpause a paused container"},
{"version", "Show the docker version information"},
{"wait", "Block until a container stops, then print its exit code"},
} {
@@ -104,11 +109,12 @@ func (cli *DockerCli) CmdInsert(args ...string) error {
}
func (cli *DockerCli) CmdBuild(args ...string) error {
cmd := cli.Subcmd("build", "[OPTIONS] PATH | URL | -", "Build a new container image from the source code at PATH")
cmd := cli.Subcmd("build", "[OPTIONS] PATH | URL | -", "Build a new image from the source code at PATH")
tag := cmd.String([]string{"t", "-tag"}, "", "Repository name (and optionally a tag) to be applied to the resulting image in case of success")
suppressOutput := cmd.Bool([]string{"q", "-quiet"}, false, "Suppress the verbose output generated by the containers")
noCache := cmd.Bool([]string{"#no-cache", "-no-cache"}, false, "Do not use cache when building the image")
rm := cmd.Bool([]string{"#rm", "-rm"}, true, "Remove intermediate containers after a successful build")
forceRm := cmd.Bool([]string{"-force-rm"}, false, "Always remove intermediate containers, even after unsuccessful builds")
if err := cmd.Parse(args); err != nil {
return nil
}
@@ -160,6 +166,9 @@ func (cli *DockerCli) CmdBuild(args ...string) error {
if _, err = os.Stat(filename); os.IsNotExist(err) {
return fmt.Errorf("no Dockerfile found in %s", cmd.Arg(0))
}
if err = utils.ValidateContextDirectory(root); err != nil {
return fmt.Errorf("Error checking context is accessible: '%s'. Please check permissions and try again.", err)
}
context, err = archive.Tar(root, archive.Uncompressed)
}
var body io.Reader
@@ -167,9 +176,9 @@ func (cli *DockerCli) CmdBuild(args ...string) error {
// FIXME: ProgressReader shouldn't be this annoying to use
if context != nil {
sf := utils.NewStreamFormatter(false)
body = utils.ProgressReader(context, 0, cli.err, sf, true, "", "Uploading context")
body = utils.ProgressReader(context, 0, cli.err, sf, true, "", "Sending build context to Docker daemon")
}
// Upload the build context
// Send the build context
v := &url.Values{}
//Check if the given image name can be resolved
@@ -193,6 +202,12 @@ func (cli *DockerCli) CmdBuild(args ...string) error {
}
if *rm {
v.Set("rm", "1")
} else {
v.Set("rm", "0")
}
if *forceRm {
v.Set("forcerm", "1")
}
cli.LoadConfigFile()
@@ -359,7 +374,7 @@ func (cli *DockerCli) CmdVersion(args ...string) error {
fmt.Fprintf(cli.out, "Client version: %s\n", dockerversion.VERSION)
}
fmt.Fprintf(cli.out, "Client API version: %s\n", api.APIVERSION)
fmt.Fprintf(cli.out, "Go version (client): %s\n", goruntime.Version())
fmt.Fprintf(cli.out, "Go version (client): %s\n", runtime.Version())
if dockerversion.GITCOMMIT != "" {
fmt.Fprintf(cli.out, "Git commit (client): %s\n", dockerversion.GITCOMMIT)
}
@@ -384,16 +399,8 @@ func (cli *DockerCli) CmdVersion(args ...string) error {
if apiVersion := remoteVersion.Get("ApiVersion"); apiVersion != "" {
fmt.Fprintf(cli.out, "Server API version: %s\n", apiVersion)
}
fmt.Fprintf(cli.out, "Git commit (server): %s\n", remoteVersion.Get("GitCommit"))
fmt.Fprintf(cli.out, "Go version (server): %s\n", remoteVersion.Get("GoVersion"))
release := utils.GetReleaseVersion()
if release != "" {
fmt.Fprintf(cli.out, "Last stable version: %s", release)
if (dockerversion.VERSION != "" || remoteVersion.Exists("Version")) && (strings.Trim(dockerversion.VERSION, "-dev") != release || strings.Trim(remoteVersion.Get("Version"), "-dev") != release) {
fmt.Fprintf(cli.out, ", please update docker")
}
fmt.Fprintf(cli.out, "\n")
}
fmt.Fprintf(cli.out, "Git commit (server): %s\n", remoteVersion.Get("GitCommit"))
return nil
}
@@ -555,10 +562,14 @@ func (cli *DockerCli) forwardAllSignals(cid string) chan os.Signal {
func (cli *DockerCli) CmdStart(args ...string) error {
var (
cErr chan error
tty bool
cmd = cli.Subcmd("start", "CONTAINER [CONTAINER...]", "Restart a stopped container")
attach = cmd.Bool([]string{"a", "-attach"}, false, "Attach container's stdout/stderr and forward all signals to the process")
openStdin = cmd.Bool([]string{"i", "-interactive"}, false, "Attach container's stdin")
)
if err := cmd.Parse(args); err != nil {
return nil
}
@@ -567,29 +578,24 @@ func (cli *DockerCli) CmdStart(args ...string) error {
return nil
}
var (
cErr chan error
tty bool
)
if *attach || *openStdin {
if cmd.NArg() > 1 {
return fmt.Errorf("You cannot start and attach multiple containers at once.")
}
body, _, err := readBody(cli.call("GET", "/containers/"+cmd.Arg(0)+"/json", nil, false))
steam, _, err := cli.call("GET", "/containers/"+cmd.Arg(0)+"/json", nil, false)
if err != nil {
return err
}
container := &api.Container{}
err = json.Unmarshal(body, container)
if err != nil {
env := engine.Env{}
if err := env.Decode(steam); err != nil {
return err
}
config := env.GetSubEnv("Config")
tty = config.GetBool("Tty")
tty = container.Config.Tty
if !container.Config.Tty {
if !tty {
sigc := cli.forwardAllSignals(cmd.Arg(0))
defer signal.StopCatch(sigc)
}
@@ -598,15 +604,17 @@ func (cli *DockerCli) CmdStart(args ...string) error {
v := url.Values{}
v.Set("stream", "1")
if *openStdin && container.Config.OpenStdin {
if *openStdin && config.GetBool("OpenStdin") {
v.Set("stdin", "1")
in = cli.in
}
v.Set("stdout", "1")
v.Set("stderr", "1")
cErr = utils.Go(func() error {
return cli.hijack("POST", "/containers/"+cmd.Arg(0)+"/attach?"+v.Encode(), container.Config.Tty, in, cli.out, cli.err, nil)
return cli.hijack("POST", "/containers/"+cmd.Arg(0)+"/attach?"+v.Encode(), tty, in, cli.out, cli.err, nil)
})
}
@@ -643,6 +651,52 @@ func (cli *DockerCli) CmdStart(args ...string) error {
return nil
}
func (cli *DockerCli) CmdUnpause(args ...string) error {
cmd := cli.Subcmd("unpause", "CONTAINER", "Unpause all processes within a container")
if err := cmd.Parse(args); err != nil {
return nil
}
if cmd.NArg() != 1 {
cmd.Usage()
return nil
}
var encounteredError error
for _, name := range cmd.Args() {
if _, _, err := readBody(cli.call("POST", fmt.Sprintf("/containers/%s/unpause", name), nil, false)); err != nil {
fmt.Fprintf(cli.err, "%s\n", err)
encounteredError = fmt.Errorf("Error: failed to unpause container named %s", name)
} else {
fmt.Fprintf(cli.out, "%s\n", name)
}
}
return encounteredError
}
func (cli *DockerCli) CmdPause(args ...string) error {
cmd := cli.Subcmd("pause", "CONTAINER", "Pause all processes within a container")
if err := cmd.Parse(args); err != nil {
return nil
}
if cmd.NArg() != 1 {
cmd.Usage()
return nil
}
var encounteredError error
for _, name := range cmd.Args() {
if _, _, err := readBody(cli.call("POST", fmt.Sprintf("/containers/%s/pause", name), nil, false)); err != nil {
fmt.Fprintf(cli.err, "%s\n", err)
encounteredError = fmt.Errorf("Error: failed to pause container named %s", name)
} else {
fmt.Fprintf(cli.out, "%s\n", name)
}
}
return encounteredError
}
func (cli *DockerCli) CmdInspect(args ...string) error {
cmd := cli.Subcmd("inspect", "CONTAINER|IMAGE [CONTAINER|IMAGE...]", "Return low-level information on a container/image")
tmplStr := cmd.String([]string{"f", "#format", "-format"}, "", "Format the output using the given go template.")
@@ -769,34 +823,38 @@ func (cli *DockerCli) CmdPort(args ...string) error {
}
var (
port = cmd.Arg(1)
proto = "tcp"
parts = strings.SplitN(port, "/", 2)
container api.Container
port = cmd.Arg(1)
proto = "tcp"
parts = strings.SplitN(port, "/", 2)
)
if len(parts) == 2 && len(parts[1]) != 0 {
port = parts[0]
proto = parts[1]
}
body, _, err := readBody(cli.call("GET", "/containers/"+cmd.Arg(0)+"/json", nil, false))
steam, _, err := cli.call("GET", "/containers/"+cmd.Arg(0)+"/json", nil, false)
if err != nil {
return err
}
err = json.Unmarshal(body, &container)
if err != nil {
env := engine.Env{}
if err := env.Decode(steam); err != nil {
return err
}
ports := nat.PortMap{}
if err := env.GetSubEnv("NetworkSettings").GetJson("Ports", &ports); err != nil {
return err
}
if frontends, exists := container.NetworkSettings.Ports[nat.Port(port+"/"+proto)]; exists && frontends != nil {
if frontends, exists := ports[nat.Port(port+"/"+proto)]; exists && frontends != nil {
for _, frontend := range frontends {
fmt.Fprintf(cli.out, "%s:%s\n", frontend.HostIp, frontend.HostPort)
}
} else {
return fmt.Errorf("Error: No public port '%s' published for %s", cmd.Arg(1), cmd.Arg(0))
return nil
}
return nil
return fmt.Errorf("Error: No public port '%s' published for %s", cmd.Arg(1), cmd.Arg(0))
}
// 'docker rmi IMAGE' removes all images with the name IMAGE
@@ -884,14 +942,14 @@ func (cli *DockerCli) CmdHistory(args ...string) error {
fmt.Fprintf(w, "%s\t", utils.TruncateID(outID))
}
fmt.Fprintf(w, "%s ago\t", utils.HumanDuration(time.Now().UTC().Sub(time.Unix(out.GetInt64("Created"), 0))))
fmt.Fprintf(w, "%s ago\t", units.HumanDuration(time.Now().UTC().Sub(time.Unix(out.GetInt64("Created"), 0))))
if *noTrunc {
fmt.Fprintf(w, "%s\t", out.Get("CreatedBy"))
} else {
fmt.Fprintf(w, "%s\t", utils.Trunc(out.Get("CreatedBy"), 45))
}
fmt.Fprintf(w, "%s\n", utils.HumanSize(out.GetInt64("Size")))
fmt.Fprintf(w, "%s\n", units.HumanSize(out.GetInt64("Size")))
} else {
if *noTrunc {
fmt.Fprintln(w, outID)
@@ -1142,6 +1200,9 @@ func (cli *DockerCli) CmdImages(args ...string) error {
flViz := cmd.Bool([]string{"#v", "#viz", "#-viz"}, false, "Output graph in graphviz format")
flTree := cmd.Bool([]string{"#t", "#tree", "#-tree"}, false, "Output graph in tree format")
var flFilter opts.ListOpts
cmd.Var(&flFilter, []string{"f", "-filter"}, "Provide filter values (i.e. 'dangling=true')")
if err := cmd.Parse(args); err != nil {
return nil
}
@@ -1150,11 +1211,32 @@ func (cli *DockerCli) CmdImages(args ...string) error {
return nil
}
filter := cmd.Arg(0)
// Consolidate all filter flags, and sanity check them early.
// They'll get process in the daemon/server.
imageFilterArgs := filters.Args{}
for _, f := range flFilter.GetAll() {
var err error
imageFilterArgs, err = filters.ParseFlag(f, imageFilterArgs)
if err != nil {
return err
}
}
matchName := cmd.Arg(0)
// FIXME: --viz and --tree are deprecated. Remove them in a future version.
if *flViz || *flTree {
body, _, err := readBody(cli.call("GET", "/images/json?all=1", nil, false))
v := url.Values{
"all": []string{"1"},
}
if len(imageFilterArgs) > 0 {
filterJson, err := filters.ToParam(imageFilterArgs)
if err != nil {
return err
}
v.Set("filters", filterJson)
}
body, _, err := readBody(cli.call("GET", "/images/json?"+v.Encode(), nil, false))
if err != nil {
return err
}
@@ -1184,13 +1266,13 @@ func (cli *DockerCli) CmdImages(args ...string) error {
}
}
if filter != "" {
if filter == image.Get("Id") || filter == utils.TruncateID(image.Get("Id")) {
if matchName != "" {
if matchName == image.Get("Id") || matchName == utils.TruncateID(image.Get("Id")) {
startImage = image
}
for _, repotag := range image.GetList("RepoTags") {
if repotag == filter {
if repotag == matchName {
startImage = image
}
}
@@ -1208,7 +1290,7 @@ func (cli *DockerCli) CmdImages(args ...string) error {
root := engine.NewTable("Created", 1)
root.Add(startImage)
cli.WalkTree(*noTrunc, root, byParent, "", printNode)
} else if filter == "" {
} else if matchName == "" {
cli.WalkTree(*noTrunc, roots, byParent, "", printNode)
}
if *flViz {
@@ -1216,8 +1298,17 @@ func (cli *DockerCli) CmdImages(args ...string) error {
}
} else {
v := url.Values{}
if len(imageFilterArgs) > 0 {
filterJson, err := filters.ToParam(imageFilterArgs)
if err != nil {
return err
}
v.Set("filters", filterJson)
}
if cmd.NArg() == 1 {
v.Set("filter", filter)
// FIXME rename this parameter, to not be confused with the filters flag
v.Set("filter", matchName)
}
if *all {
v.Set("all", "1")
@@ -1249,7 +1340,7 @@ func (cli *DockerCli) CmdImages(args ...string) error {
}
if !*quiet {
fmt.Fprintf(w, "%s\t%s\t%s\t%s ago\t%s\n", repo, tag, outID, utils.HumanDuration(time.Now().UTC().Sub(time.Unix(out.GetInt64("Created"), 0))), utils.HumanSize(out.GetInt64("VirtualSize")))
fmt.Fprintf(w, "%s\t%s\t%s\t%s ago\t%s\n", repo, tag, outID, units.HumanDuration(time.Now().UTC().Sub(time.Unix(out.GetInt64("Created"), 0))), units.HumanSize(out.GetInt64("VirtualSize")))
} else {
fmt.Fprintln(w, outID)
}
@@ -1323,7 +1414,7 @@ func (cli *DockerCli) printTreeNode(noTrunc bool, image *engine.Env, prefix stri
imageID = utils.TruncateID(image.Get("Id"))
}
fmt.Fprintf(cli.out, "%s%s Virtual Size: %s", prefix, imageID, utils.HumanSize(image.GetInt64("VirtualSize")))
fmt.Fprintf(cli.out, "%s%s Virtual Size: %s", prefix, imageID, units.HumanSize(image.GetInt64("VirtualSize")))
if image.GetList("RepoTags")[0] != "<none>:<none>" {
fmt.Fprintf(cli.out, " Tags: %s\n", strings.Join(image.GetList("RepoTags"), ", "))
} else {
@@ -1408,12 +1499,12 @@ func (cli *DockerCli) CmdPs(args ...string) error {
outCommand = utils.Trunc(outCommand, 20)
}
ports.ReadListFrom([]byte(out.Get("Ports")))
fmt.Fprintf(w, "%s\t%s\t%s\t%s ago\t%s\t%s\t%s\t", outID, out.Get("Image"), outCommand, utils.HumanDuration(time.Now().UTC().Sub(time.Unix(out.GetInt64("Created"), 0))), out.Get("Status"), api.DisplayablePorts(ports), strings.Join(outNames, ","))
fmt.Fprintf(w, "%s\t%s\t%s\t%s ago\t%s\t%s\t%s\t", outID, out.Get("Image"), outCommand, units.HumanDuration(time.Now().UTC().Sub(time.Unix(out.GetInt64("Created"), 0))), out.Get("Status"), api.DisplayablePorts(ports), strings.Join(outNames, ","))
if *size {
if out.GetInt("SizeRootFs") > 0 {
fmt.Fprintf(w, "%s (virtual %s)\n", utils.HumanSize(out.GetInt64("SizeRw")), utils.HumanSize(out.GetInt64("SizeRootFs")))
fmt.Fprintf(w, "%s (virtual %s)\n", units.HumanSize(out.GetInt64("SizeRw")), units.HumanSize(out.GetInt64("SizeRootFs")))
} else {
fmt.Fprintf(w, "%s\n", utils.HumanSize(out.GetInt64("SizeRw")))
fmt.Fprintf(w, "%s\n", units.HumanSize(out.GetInt64("SizeRw")))
}
} else {
fmt.Fprint(w, "\n")
@@ -1581,72 +1672,84 @@ func (cli *DockerCli) CmdDiff(args ...string) error {
}
func (cli *DockerCli) CmdLogs(args ...string) error {
cmd := cli.Subcmd("logs", "CONTAINER", "Fetch the logs of a container")
follow := cmd.Bool([]string{"f", "-follow"}, false, "Follow log output")
times := cmd.Bool([]string{"t", "-timestamps"}, false, "Show timestamps")
var (
cmd = cli.Subcmd("logs", "CONTAINER", "Fetch the logs of a container")
follow = cmd.Bool([]string{"f", "-follow"}, false, "Follow log output")
times = cmd.Bool([]string{"t", "-timestamps"}, false, "Show timestamps")
)
if err := cmd.Parse(args); err != nil {
return nil
}
if cmd.NArg() != 1 {
cmd.Usage()
return nil
}
name := cmd.Arg(0)
body, _, err := readBody(cli.call("GET", "/containers/"+name+"/json", nil, false))
steam, _, err := cli.call("GET", "/containers/"+name+"/json", nil, false)
if err != nil {
return err
}
container := &api.Container{}
err = json.Unmarshal(body, container)
if err != nil {
env := engine.Env{}
if err := env.Decode(steam); err != nil {
return err
}
v := url.Values{}
v.Set("stdout", "1")
v.Set("stderr", "1")
if *times {
v.Set("timestamps", "1")
}
if *follow && container.State.Running {
if *follow {
v.Set("follow", "1")
}
if err := cli.streamHelper("GET", "/containers/"+name+"/logs?"+v.Encode(), container.Config.Tty, nil, cli.out, cli.err, nil); err != nil {
return err
}
return nil
return cli.streamHelper("GET", "/containers/"+name+"/logs?"+v.Encode(), env.GetSubEnv("Config").GetBool("Tty"), nil, cli.out, cli.err, nil)
}
func (cli *DockerCli) CmdAttach(args ...string) error {
cmd := cli.Subcmd("attach", "[OPTIONS] CONTAINER", "Attach to a running container")
noStdin := cmd.Bool([]string{"#nostdin", "-no-stdin"}, false, "Do not attach stdin")
proxy := cmd.Bool([]string{"#sig-proxy", "-sig-proxy"}, true, "Proxify all received signal to the process (even in non-tty mode)")
var (
cmd = cli.Subcmd("attach", "[OPTIONS] CONTAINER", "Attach to a running container")
noStdin = cmd.Bool([]string{"#nostdin", "-no-stdin"}, false, "Do not attach stdin")
proxy = cmd.Bool([]string{"#sig-proxy", "-sig-proxy"}, true, "Proxify all received signal to the process (even in non-tty mode)")
)
if err := cmd.Parse(args); err != nil {
return nil
}
if cmd.NArg() != 1 {
cmd.Usage()
return nil
}
name := cmd.Arg(0)
body, _, err := readBody(cli.call("GET", "/containers/"+name+"/json", nil, false))
stream, _, err := cli.call("GET", "/containers/"+name+"/json", nil, false)
if err != nil {
return err
}
container := &api.Container{}
err = json.Unmarshal(body, container)
if err != nil {
env := engine.Env{}
if err := env.Decode(stream); err != nil {
return err
}
if !container.State.Running {
if !env.GetSubEnv("State").GetBool("Running") {
return fmt.Errorf("You cannot attach to a stopped container, start it first")
}
if container.Config.Tty && cli.isTerminal {
var (
config = env.GetSubEnv("Config")
tty = config.GetBool("Tty")
)
if tty && cli.isTerminal {
if err := cli.monitorTtySize(cmd.Arg(0)); err != nil {
utils.Debugf("Error monitoring TTY size: %s", err)
}
@@ -1656,19 +1759,20 @@ func (cli *DockerCli) CmdAttach(args ...string) error {
v := url.Values{}
v.Set("stream", "1")
if !*noStdin && container.Config.OpenStdin {
if !*noStdin && config.GetBool("OpenStdin") {
v.Set("stdin", "1")
in = cli.in
}
v.Set("stdout", "1")
v.Set("stderr", "1")
if *proxy && !container.Config.Tty {
if *proxy && !tty {
sigc := cli.forwardAllSignals(cmd.Arg(0))
defer signal.StopCatch(sigc)
}
if err := cli.hijack("POST", "/containers/"+cmd.Arg(0)+"/attach?"+v.Encode(), container.Config.Tty, in, cli.out, cli.err, nil); err != nil {
if err := cli.hijack("POST", "/containers/"+cmd.Arg(0)+"/attach?"+v.Encode(), tty, in, cli.out, cli.err, nil); err != nil {
return err
}
@@ -1686,7 +1790,8 @@ func (cli *DockerCli) CmdAttach(args ...string) error {
func (cli *DockerCli) CmdSearch(args ...string) error {
cmd := cli.Subcmd("search", "TERM", "Search the docker index for images")
noTrunc := cmd.Bool([]string{"#notrunc", "-no-trunc"}, false, "Don't truncate output")
trusted := cmd.Bool([]string{"t", "#trusted", "-trusted"}, false, "Only show trusted builds")
trusted := cmd.Bool([]string{"#t", "#trusted", "#-trusted"}, false, "Only show trusted builds")
automated := cmd.Bool([]string{"-automated"}, false, "Only show automated builds")
stars := cmd.Int([]string{"s", "#stars", "-stars"}, 0, "Only displays with at least xxx stars")
if err := cmd.Parse(args); err != nil {
return nil
@@ -1709,9 +1814,9 @@ func (cli *DockerCli) CmdSearch(args ...string) error {
return err
}
w := tabwriter.NewWriter(cli.out, 10, 1, 3, ' ', 0)
fmt.Fprintf(w, "NAME\tDESCRIPTION\tSTARS\tOFFICIAL\tTRUSTED\n")
fmt.Fprintf(w, "NAME\tDESCRIPTION\tSTARS\tOFFICIAL\tAUTOMATED\n")
for _, out := range outs.Data {
if (*trusted && !out.GetBool("is_trusted")) || (*stars > out.GetInt("star_count")) {
if ((*automated || *trusted) && (!out.GetBool("is_trusted") && !out.GetBool("is_automated"))) || (*stars > out.GetInt("star_count")) {
continue
}
desc := strings.Replace(out.Get("description"), "\n", " ", -1)
@@ -1725,7 +1830,7 @@ func (cli *DockerCli) CmdSearch(args ...string) error {
}
fmt.Fprint(w, "\t")
if out.GetBool("is_trusted") {
if out.GetBool("is_automated") || out.GetBool("is_trusted") {
fmt.Fprint(w, "[OK]")
}
fmt.Fprint(w, "\n")
@@ -1839,6 +1944,10 @@ func (cli *DockerCli) CmdRun(args ...string) error {
v := url.Values{}
repos, tag := utils.ParseRepositoryTag(config.Image)
// pull only the image tagged 'latest' if no tag was specified
if tag == "" {
tag = "latest"
}
v.Set("fromImage", repos)
v.Set("tag", tag)
@@ -2058,7 +2167,7 @@ func (cli *DockerCli) CmdCp(args ...string) error {
}
if statusCode == 200 {
if err := archive.Untar(stream, copyData.Get("HostPath"), nil); err != nil {
if err := archive.Untar(stream, copyData.Get("HostPath"), &archive.TarOptions{NoLchown: true}); err != nil {
return err
}
}
+7 -4
View File
@@ -200,7 +200,7 @@ func waitForExit(cli *DockerCli, containerId string) (int, error) {
// getExitCode perform an inspect on the container. It returns
// the running state and the exit code.
func getExitCode(cli *DockerCli, containerId string) (bool, int, error) {
body, _, err := readBody(cli.call("GET", "/containers/"+containerId+"/json", nil, false))
steam, _, err := cli.call("GET", "/containers/"+containerId+"/json", nil, false)
if err != nil {
// If we can't connect, then the daemon probably died.
if err != ErrConnectionRefused {
@@ -208,11 +208,14 @@ func getExitCode(cli *DockerCli, containerId string) (bool, int, error) {
}
return false, -1, nil
}
c := &api.Container{}
if err := json.Unmarshal(body, c); err != nil {
var result engine.Env
if err := result.Decode(steam); err != nil {
return false, -1, err
}
return c.State.Running, c.State.ExitCode, nil
state := result.GetSubEnv("State")
return state.GetBool("Running"), state.GetInt("ExitCode"), nil
}
func (cli *DockerCli) monitorTtySize(id string) error {
+5 -4
View File
@@ -2,15 +2,16 @@ package api
import (
"fmt"
"mime"
"strings"
"github.com/dotcloud/docker/engine"
"github.com/dotcloud/docker/pkg/version"
"github.com/dotcloud/docker/utils"
"mime"
"strings"
)
const (
APIVERSION version.Version = "1.11"
APIVERSION version.Version = "1.12"
DEFAULTHTTPHOST = "127.0.0.1"
DEFAULTUNIXSOCKET = "/var/run/docker.sock"
)
@@ -30,7 +31,7 @@ func DisplayablePorts(ports *engine.Table) string {
ports.Sort()
for _, port := range ports.Data {
if port.Get("IP") == "" {
result = append(result, fmt.Sprintf("%d/%s", port.GetInt("PublicPort"), port.Get("Type")))
result = append(result, fmt.Sprintf("%d/%s", port.GetInt("PrivatePort"), port.Get("Type")))
} else {
result = append(result, fmt.Sprintf("%s:%d->%d/%s", port.Get("IP"), port.GetInt("PublicPort"), port.GetInt("PrivatePort"), port.Get("Type")))
}
-18
View File
@@ -1,18 +0,0 @@
package api
import (
"github.com/dotcloud/docker/nat"
"github.com/dotcloud/docker/runconfig"
)
type Container struct {
Config runconfig.Config
HostConfig runconfig.HostConfig
State struct {
Running bool
ExitCode int
}
NetworkSettings struct {
Ports nat.PortMap
}
}
+98 -71
View File
@@ -122,17 +122,17 @@ func postAuth(eng *engine.Engine, version version.Version, w http.ResponseWriter
var (
authConfig, err = ioutil.ReadAll(r.Body)
job = eng.Job("auth")
status string
stdoutBuffer = bytes.NewBuffer(nil)
)
if err != nil {
return err
}
job.Setenv("authConfig", string(authConfig))
job.Stdout.AddString(&status)
job.Stdout.Add(stdoutBuffer)
if err = job.Run(); err != nil {
return err
}
if status != "" {
if status := engine.Tail(stdoutBuffer, 1); status != "" {
var env engine.Env
env.Set("Status", status)
return writeJSON(w, http.StatusOK, env)
@@ -165,6 +165,36 @@ func postContainersKill(eng *engine.Engine, version version.Version, w http.Resp
return nil
}
func postContainersPause(eng *engine.Engine, version version.Version, w http.ResponseWriter, r *http.Request, vars map[string]string) error {
if vars == nil {
return fmt.Errorf("Missing parameter")
}
if err := parseForm(r); err != nil {
return err
}
job := eng.Job("pause", vars["name"])
if err := job.Run(); err != nil {
return err
}
w.WriteHeader(http.StatusNoContent)
return nil
}
func postContainersUnpause(eng *engine.Engine, version version.Version, w http.ResponseWriter, r *http.Request, vars map[string]string) error {
if vars == nil {
return fmt.Errorf("Missing parameter")
}
if err := parseForm(r); err != nil {
return err
}
job := eng.Job("unpause", vars["name"])
if err := job.Run(); err != nil {
return err
}
w.WriteHeader(http.StatusNoContent)
return nil
}
func getContainersExport(eng *engine.Engine, version version.Version, w http.ResponseWriter, r *http.Request, vars map[string]string) error {
if vars == nil {
return fmt.Errorf("Missing parameter")
@@ -188,6 +218,8 @@ func getImagesJSON(eng *engine.Engine, version version.Version, w http.ResponseW
job = eng.Job("images")
)
job.Setenv("filters", r.Form.Get("filters"))
// FIXME this parameter could just be a match filter
job.Setenv("filter", r.Form.Get("filter"))
job.Setenv("all", r.Form.Get("all"))
@@ -244,7 +276,7 @@ func getEvents(eng *engine.Engine, version version.Version, w http.ResponseWrite
return err
}
var job = eng.Job("events", r.RemoteAddr)
var job = eng.Job("events")
streamJSON(job, w, true)
job.Setenv("since", r.Form.Get("since"))
job.Setenv("until", r.Form.Get("until"))
@@ -338,7 +370,7 @@ func getContainersLogs(eng *engine.Engine, version version.Version, w http.Respo
}
var (
job = eng.Job("inspect", vars["name"], "container")
job = eng.Job("container_inspect", vars["name"])
c, err = job.Stdout.AddEnv()
)
if err != nil {
@@ -393,9 +425,10 @@ func postCommit(eng *engine.Engine, version version.Version, w http.ResponseWrit
return err
}
var (
config engine.Env
env engine.Env
job = eng.Job("commit", r.Form.Get("container"))
config engine.Env
env engine.Env
job = eng.Job("commit", r.Form.Get("container"))
stdoutBuffer = bytes.NewBuffer(nil)
)
if err := config.Decode(r.Body); err != nil {
utils.Errorf("%s", err)
@@ -407,12 +440,11 @@ func postCommit(eng *engine.Engine, version version.Version, w http.ResponseWrit
job.Setenv("comment", r.Form.Get("comment"))
job.SetenvSubEnv("config", &config)
var id string
job.Stdout.AddString(&id)
job.Stdout.Add(stdoutBuffer)
if err := job.Run(); err != nil {
return err
}
env.Set("Id", id)
env.Set("Id", engine.Tail(stdoutBuffer, 1))
return writeJSON(w, http.StatusCreated, env)
}
@@ -502,32 +534,6 @@ func getImagesSearch(eng *engine.Engine, version version.Version, w http.Respons
return job.Run()
}
// FIXME: 'insert' is deprecated as of 0.10, and should be removed in a future version.
func postImagesInsert(eng *engine.Engine, version version.Version, w http.ResponseWriter, r *http.Request, vars map[string]string) error {
if err := parseForm(r); err != nil {
return err
}
if vars == nil {
return fmt.Errorf("Missing parameter")
}
job := eng.Job("insert", vars["name"], r.Form.Get("url"), r.Form.Get("path"))
if version.GreaterThan("1.0") {
job.SetenvBool("json", true)
streamJSON(job, w, false)
} else {
job.Stdout.Add(w)
}
if err := job.Run(); err != nil {
if !job.Stdout.Used() {
return err
}
sf := utils.NewStreamFormatter(version.GreaterThan("1.0"))
w.Write(sf.FormatError(err))
}
return nil
}
func postImagesPush(eng *engine.Engine, version version.Version, w http.ResponseWriter, r *http.Request, vars map[string]string) error {
if vars == nil {
return fmt.Errorf("Missing parameter")
@@ -603,17 +609,17 @@ func postContainersCreate(eng *engine.Engine, version version.Version, w http.Re
return nil
}
var (
out engine.Env
job = eng.Job("create", r.Form.Get("name"))
outWarnings []string
outId string
warnings = bytes.NewBuffer(nil)
out engine.Env
job = eng.Job("create", r.Form.Get("name"))
outWarnings []string
stdoutBuffer = bytes.NewBuffer(nil)
warnings = bytes.NewBuffer(nil)
)
if err := job.DecodeEnv(r.Body); err != nil {
return err
}
// Read container ID from the first line of stdout
job.Stdout.AddString(&outId)
job.Stdout.Add(stdoutBuffer)
// Read warnings from stderr
job.Stderr.Add(warnings)
if err := job.Run(); err != nil {
@@ -624,7 +630,7 @@ func postContainersCreate(eng *engine.Engine, version version.Version, w http.Re
for scanner.Scan() {
outWarnings = append(outWarnings, scanner.Text())
}
out.Set("Id", outId)
out.Set("Id", engine.Tail(stdoutBuffer, 1))
out.SetList("Warnings", outWarnings)
return writeJSON(w, http.StatusCreated, out)
}
@@ -720,20 +726,16 @@ func postContainersWait(eng *engine.Engine, version version.Version, w http.Resp
return fmt.Errorf("Missing parameter")
}
var (
env engine.Env
status string
job = eng.Job("wait", vars["name"])
env engine.Env
stdoutBuffer = bytes.NewBuffer(nil)
job = eng.Job("wait", vars["name"])
)
job.Stdout.AddString(&status)
job.Stdout.Add(stdoutBuffer)
if err := job.Run(); err != nil {
return err
}
// Parse a 16-bit encoded integer to map typical unix exit status.
_, err := strconv.ParseInt(status, 10, 16)
if err != nil {
return err
}
env.Set("StatusCode", status)
env.Set("StatusCode", engine.Tail(stdoutBuffer, 1))
return writeJSON(w, http.StatusOK, env)
}
@@ -759,7 +761,7 @@ func postContainersAttach(eng *engine.Engine, version version.Version, w http.Re
}
var (
job = eng.Job("inspect", vars["name"], "container")
job = eng.Job("container_inspect", vars["name"])
c, err = job.Stdout.AddEnv()
)
if err != nil {
@@ -823,7 +825,7 @@ func wsContainersAttach(eng *engine.Engine, version version.Version, w http.Resp
return fmt.Errorf("Missing parameter")
}
if err := eng.Job("inspect", vars["name"], "container").Run(); err != nil {
if err := eng.Job("container_inspect", vars["name"]).Run(); err != nil {
return err
}
@@ -851,9 +853,11 @@ func getContainersByName(eng *engine.Engine, version version.Version, w http.Res
if vars == nil {
return fmt.Errorf("Missing parameter")
}
var job = eng.Job("inspect", vars["name"], "container")
var job = eng.Job("container_inspect", vars["name"])
if version.LessThan("1.12") {
job.SetenvBool("dirty", true)
}
streamJSON(job, w, false)
job.SetenvBool("conflict", true) //conflict=true to detect conflict between containers and images in the job
return job.Run()
}
@@ -861,9 +865,11 @@ func getImagesByName(eng *engine.Engine, version version.Version, w http.Respons
if vars == nil {
return fmt.Errorf("Missing parameter")
}
var job = eng.Job("inspect", vars["name"], "image")
var job = eng.Job("image_inspect", vars["name"])
if version.LessThan("1.12") {
job.SetenvBool("dirty", true)
}
streamJSON(job, w, false)
job.SetenvBool("conflict", true) //conflict=true to detect conflict between containers and images in the job
return job.Run()
}
@@ -907,12 +913,20 @@ func postBuild(eng *engine.Engine, version version.Version, w http.ResponseWrite
} else {
job.Stdout.Add(utils.NewWriteFlusher(w))
}
if r.FormValue("forcerm") == "1" && version.GreaterThanOrEqualTo("1.12") {
job.Setenv("rm", "1")
} else if r.FormValue("rm") == "" && version.GreaterThanOrEqualTo("1.12") {
job.Setenv("rm", "1")
} else {
job.Setenv("rm", r.FormValue("rm"))
}
job.Stdin.Add(r.Body)
job.Setenv("remote", r.FormValue("remote"))
job.Setenv("t", r.FormValue("t"))
job.Setenv("q", r.FormValue("q"))
job.Setenv("nocache", r.FormValue("nocache"))
job.Setenv("rm", r.FormValue("rm"))
job.Setenv("forcerm", r.FormValue("forcerm"))
job.SetenvJson("authConfig", authConfig)
job.SetenvJson("configFile", configFile)
@@ -1071,12 +1085,13 @@ func createRouter(eng *engine.Engine, logging, enableCors bool, dockerVersion st
"/commit": postCommit,
"/build": postBuild,
"/images/create": postImagesCreate,
"/images/{name:.*}/insert": postImagesInsert,
"/images/load": postImagesLoad,
"/images/{name:.*}/push": postImagesPush,
"/images/{name:.*}/tag": postImagesTag,
"/containers/create": postContainersCreate,
"/containers/{name:.*}/kill": postContainersKill,
"/containers/{name:.*}/pause": postContainersPause,
"/containers/{name:.*}/unpause": postContainersUnpause,
"/containers/{name:.*}/restart": postContainersRestart,
"/containers/{name:.*}/start": postContainersStart,
"/containers/{name:.*}/stop": postContainersStop,
@@ -1193,6 +1208,7 @@ func changeGroup(addr string, nameOrGid string) error {
// ListenAndServe sets up the required http.Server and gets it listening for
// each addr passed in and does protocol specific checking.
func ListenAndServe(proto, addr string, job *engine.Job) error {
var l net.Listener
r, err := createRouter(job.Eng, job.GetenvBool("Logging"), job.GetenvBool("EnableCors"), job.Getenv("Version"))
if err != nil {
return err
@@ -1208,7 +1224,20 @@ func ListenAndServe(proto, addr string, job *engine.Job) error {
}
}
l, err := listenbuffer.NewListenBuffer(proto, addr, activationLock)
var oldmask int
if proto == "unix" {
oldmask = syscall.Umask(0777)
}
if job.GetenvBool("BufferRequests") {
l, err = listenbuffer.NewListenBuffer(proto, addr, activationLock)
} else {
l, err = net.Listen(proto, addr)
}
if proto == "unix" {
syscall.Umask(oldmask)
}
if err != nil {
return err
}
@@ -1246,9 +1275,6 @@ func ListenAndServe(proto, addr string, job *engine.Job) error {
log.Println("/!\\ DON'T BIND ON ANOTHER IP ADDRESS THAN 127.0.0.1 IF YOU DON'T KNOW WHAT YOU'RE DOING /!\\")
}
case "unix":
if err := os.Chmod(addr, 0660); err != nil {
return err
}
socketGroup := job.Getenv("SocketGroup")
if socketGroup != "" {
if err := changeGroup(addr, socketGroup); err != nil {
@@ -1260,6 +1286,9 @@ func ListenAndServe(proto, addr string, job *engine.Job) error {
}
}
}
if err := os.Chmod(addr, 0660); err != nil {
return err
}
default:
return fmt.Errorf("Invalid protocol format.")
}
@@ -1280,10 +1309,6 @@ func ServeApi(job *engine.Job) engine.Status {
)
activationLock = make(chan struct{})
if err := job.Eng.Register("acceptconnections", AcceptConnections); err != nil {
return job.Error(err)
}
for _, protoAddr := range protoAddrs {
protoAddrParts := strings.SplitN(protoAddr, "://", 2)
if len(protoAddrParts) != 2 {
@@ -1310,7 +1335,9 @@ func AcceptConnections(job *engine.Job) engine.Status {
go systemd.SdNotify("READY=1")
// close the lock so the listeners start accepting connections
close(activationLock)
if activationLock != nil {
close(activationLock)
}
return engine.StatusOK
}
+3
View File
@@ -0,0 +1,3 @@
This code provides helper functions for dealing with archive files.
**TODO**: Move this to either `pkg` or (if not possible) to `utils`.
+30 -39
View File
@@ -1,14 +1,12 @@
package archive
import (
"bufio"
"bytes"
"compress/bzip2"
"compress/gzip"
"errors"
"fmt"
"github.com/dotcloud/docker/pkg/system"
"github.com/dotcloud/docker/utils"
"github.com/dotcloud/docker/vendor/src/code.google.com/p/go/src/pkg/archive/tar"
"io"
"io/ioutil"
"os"
@@ -17,6 +15,10 @@ import (
"path/filepath"
"strings"
"syscall"
"github.com/dotcloud/docker/pkg/system"
"github.com/dotcloud/docker/utils"
"github.com/dotcloud/docker/vendor/src/code.google.com/p/go/src/pkg/archive/tar"
)
type (
@@ -26,6 +28,7 @@ type (
TarOptions struct {
Includes []string
Compression Compression
NoLchown bool
}
)
@@ -41,26 +44,16 @@ const (
)
func DetectCompression(source []byte) Compression {
sourceLen := len(source)
for compression, m := range map[Compression][]byte{
Bzip2: {0x42, 0x5A, 0x68},
Gzip: {0x1F, 0x8B, 0x08},
Xz: {0xFD, 0x37, 0x7A, 0x58, 0x5A, 0x00},
} {
fail := false
if len(m) > sourceLen {
if len(source) < len(m) {
utils.Debugf("Len too short")
continue
}
i := 0
for _, b := range m {
if b != source[i] {
fail = true
break
}
i++
}
if !fail {
if bytes.Compare(m, source[:len(m)]) == 0 {
return compression
}
}
@@ -74,31 +67,24 @@ func xzDecompress(archive io.Reader) (io.ReadCloser, error) {
}
func DecompressStream(archive io.Reader) (io.ReadCloser, error) {
buf := make([]byte, 10)
totalN := 0
for totalN < 10 {
n, err := archive.Read(buf[totalN:])
if err != nil {
if err == io.EOF {
return nil, fmt.Errorf("Tarball too short")
}
return nil, err
}
totalN += n
utils.Debugf("[tar autodetect] n: %d", n)
buf := bufio.NewReader(archive)
bs, err := buf.Peek(10)
if err != nil {
return nil, err
}
compression := DetectCompression(buf)
wrap := io.MultiReader(bytes.NewReader(buf), archive)
utils.Debugf("[tar autodetect] n: %v", bs)
compression := DetectCompression(bs)
switch compression {
case Uncompressed:
return ioutil.NopCloser(wrap), nil
return ioutil.NopCloser(buf), nil
case Gzip:
return gzip.NewReader(wrap)
return gzip.NewReader(buf)
case Bzip2:
return ioutil.NopCloser(bzip2.NewReader(wrap)), nil
return ioutil.NopCloser(bzip2.NewReader(buf)), nil
case Xz:
return xzDecompress(wrap)
return xzDecompress(buf)
default:
return nil, fmt.Errorf("Unsupported compression format %s", (&compression).Extension())
}
@@ -194,7 +180,7 @@ func addTarFile(path, name string, tw *tar.Writer) error {
return nil
}
func createTarFile(path, extractDir string, hdr *tar.Header, reader io.Reader) error {
func createTarFile(path, extractDir string, hdr *tar.Header, reader io.Reader, Lchown bool) error {
// hdr.Mode is in linux format, which we can use for sycalls,
// but for os.Foo() calls we need the mode converted to os.FileMode,
// so use hdrInfo.Mode() (they differ for e.g. setuid bits)
@@ -255,7 +241,7 @@ func createTarFile(path, extractDir string, hdr *tar.Header, reader io.Reader) e
return fmt.Errorf("Unhandled tar header type %d\n", hdr.Typeflag)
}
if err := os.Lchown(path, hdr.Uid, hdr.Gid); err != nil {
if err := os.Lchown(path, hdr.Uid, hdr.Gid); err != nil && Lchown {
return err
}
@@ -309,8 +295,11 @@ func escapeName(name string) string {
return string(escaped)
}
// Tar creates an archive from the directory at `path`, only including files whose relative
// paths are included in `filter`. If `filter` is nil, then all files are included.
// TarFilter creates an archive from the directory at `srcPath` with `options`, and returns it as a
// stream of bytes.
//
// Files are included according to `options.Includes`, default to including all files.
// Stream is compressed according to `options.Compression', default to Uncompressed.
func TarFilter(srcPath string, options *TarOptions) (io.ReadCloser, error) {
pipeReader, pipeWriter := io.Pipe()
@@ -418,14 +407,16 @@ func Untar(archive io.Reader, dest string, options *TarOptions) error {
// the layer is also a directory. Then we want to merge them (i.e.
// just apply the metadata from the layer).
if fi, err := os.Lstat(path); err == nil {
if fi.IsDir() && hdr.Name == "." {
continue
}
if !(fi.IsDir() && hdr.Typeflag == tar.TypeDir) {
if err := os.RemoveAll(path); err != nil {
return err
}
}
}
if err := createTarFile(path, dest, hdr, tr); err != nil {
if err := createTarFile(path, dest, hdr, tr, options == nil || !options.NoLchown); err != nil {
return err
}
+32 -2
View File
@@ -3,7 +3,6 @@ package archive
import (
"bytes"
"fmt"
"github.com/dotcloud/docker/vendor/src/code.google.com/p/go/src/pkg/archive/tar"
"io"
"io/ioutil"
"os"
@@ -11,6 +10,8 @@ import (
"path"
"testing"
"time"
"github.com/dotcloud/docker/vendor/src/code.google.com/p/go/src/pkg/archive/tar"
)
func TestCmdStreamLargeStderr(t *testing.T) {
@@ -132,8 +133,37 @@ func TestTarUntar(t *testing.T) {
// Failing prevents the archives from being uncompressed during ADD
func TestTypeXGlobalHeaderDoesNotFail(t *testing.T) {
hdr := tar.Header{Typeflag: tar.TypeXGlobalHeader}
err := createTarFile("pax_global_header", "some_dir", &hdr, nil)
err := createTarFile("pax_global_header", "some_dir", &hdr, nil, true)
if err != nil {
t.Fatal(err)
}
}
// Some tar have both GNU specific (huge uid) and Ustar specific (long name) things.
// Not supposed to happen (should use PAX instead of Ustar for long name) but it does and it should still work.
func TestUntarUstarGnuConflict(t *testing.T) {
f, err := os.Open("testdata/broken.tar")
if err != nil {
t.Fatal(err)
}
found := false
tr := tar.NewReader(f)
// Iterate through the files in the archive.
for {
hdr, err := tr.Next()
if err == io.EOF {
// end of tar archive
break
}
if err != nil {
t.Fatal(err)
}
if hdr.Name == "root/.cpanm/work/1395823785.24209/Plack-1.0030/blib/man3/Plack::Middleware::LighttpdScriptNameFix.3pm" {
found = true
break
}
}
if !found {
t.Fatal("%s not found in the archive", "root/.cpanm/work/1395823785.24209/Plack-1.0030/blib/man3/Plack::Middleware::LighttpdScriptNameFix.3pm")
}
}
+25 -13
View File
@@ -3,15 +3,16 @@ package archive
import (
"bytes"
"fmt"
"github.com/dotcloud/docker/pkg/system"
"github.com/dotcloud/docker/utils"
"github.com/dotcloud/docker/vendor/src/code.google.com/p/go/src/pkg/archive/tar"
"io"
"os"
"path/filepath"
"strings"
"syscall"
"time"
"github.com/dotcloud/docker/pkg/system"
"github.com/dotcloud/docker/utils"
"github.com/dotcloud/docker/vendor/src/code.google.com/p/go/src/pkg/archive/tar"
)
type ChangeType int
@@ -293,13 +294,23 @@ func collectFileInfo(sourceDir string) (*FileInfo, error) {
// Compare two directories and generate an array of Change objects describing the changes
func ChangesDirs(newDir, oldDir string) ([]Change, error) {
oldRoot, err := collectFileInfo(oldDir)
if err != nil {
return nil, err
}
newRoot, err := collectFileInfo(newDir)
if err != nil {
return nil, err
var (
oldRoot, newRoot *FileInfo
err1, err2 error
errs = make(chan error, 2)
)
go func() {
oldRoot, err1 = collectFileInfo(oldDir)
errs <- err1
}()
go func() {
newRoot, err2 = collectFileInfo(newDir)
errs <- err2
}()
for i := 0; i < 2; i++ {
if err := <-errs; err != nil {
return nil, err
}
}
return newRoot.Changes(oldRoot), nil
@@ -341,12 +352,13 @@ func ExportChanges(dir string, changes []Change) (Archive, error) {
whiteOutDir := filepath.Dir(change.Path)
whiteOutBase := filepath.Base(change.Path)
whiteOut := filepath.Join(whiteOutDir, ".wh."+whiteOutBase)
timestamp := time.Now()
hdr := &tar.Header{
Name: whiteOut[1:],
Size: 0,
ModTime: time.Now(),
AccessTime: time.Now(),
ChangeTime: time.Now(),
ModTime: timestamp,
AccessTime: timestamp,
ChangeTime: timestamp,
}
if err := tw.WriteHeader(hdr); err != nil {
utils.Debugf("Can't write whiteout header: %s\n", err)
+4 -13
View File
@@ -2,14 +2,14 @@ package archive
import (
"fmt"
"github.com/dotcloud/docker/vendor/src/code.google.com/p/go/src/pkg/archive/tar"
"io"
"io/ioutil"
"os"
"path/filepath"
"strings"
"syscall"
"time"
"github.com/dotcloud/docker/vendor/src/code.google.com/p/go/src/pkg/archive/tar"
)
// Linux device nodes are a bit weird due to backwards compat with 16 bit device nodes.
@@ -18,15 +18,6 @@ import (
func mkdev(major int64, minor int64) uint32 {
return uint32(((minor & 0xfff00) << 12) | ((major & 0xfff) << 8) | (minor & 0xff))
}
func timeToTimespec(time time.Time) (ts syscall.Timespec) {
if time.IsZero() {
// Return UTIME_OMIT special value
ts.Sec = 0
ts.Nsec = ((1 << 30) - 2)
return
}
return syscall.NsecToTimespec(time.UnixNano())
}
// ApplyLayer parses a diff in the standard layer format from `layer`, and
// applies it to the directory `dest`.
@@ -89,7 +80,7 @@ func ApplyLayer(dest string, layer ArchiveReader) error {
}
defer os.RemoveAll(aufsTempdir)
}
if err := createTarFile(filepath.Join(aufsTempdir, basename), dest, hdr, tr); err != nil {
if err := createTarFile(filepath.Join(aufsTempdir, basename), dest, hdr, tr, true); err != nil {
return err
}
}
@@ -136,7 +127,7 @@ func ApplyLayer(dest string, layer ArchiveReader) error {
srcData = tmpFile
}
if err := createTarFile(path, dest, srcHdr, srcData); err != nil {
if err := createTarFile(path, dest, srcHdr, srcData, true); err != nil {
return err
}
BIN
View File
Binary file not shown.
+16
View File
@@ -0,0 +1,16 @@
package archive
import (
"syscall"
"time"
)
func timeToTimespec(time time.Time) (ts syscall.Timespec) {
if time.IsZero() {
// Return UTIME_OMIT special value
ts.Sec = 0
ts.Nsec = ((1 << 30) - 2)
return
}
return syscall.NsecToTimespec(time.UnixNano())
}
+16
View File
@@ -0,0 +1,16 @@
// +build !linux
package archive
import (
"syscall"
"time"
)
func timeToTimespec(time time.Time) (ts syscall.Timespec) {
nsec := int64(0)
if !time.IsZero() {
nsec = time.UnixNano()
}
return syscall.NsecToTimespec(nsec)
}
+31 -2
View File
@@ -1,11 +1,16 @@
package builtins
import (
api "github.com/dotcloud/docker/api/server"
"runtime"
"github.com/dotcloud/docker/api"
apiserver "github.com/dotcloud/docker/api/server"
"github.com/dotcloud/docker/daemon/networkdriver/bridge"
"github.com/dotcloud/docker/dockerversion"
"github.com/dotcloud/docker/engine"
"github.com/dotcloud/docker/registry"
"github.com/dotcloud/docker/server"
"github.com/dotcloud/docker/utils"
)
func Register(eng *engine.Engine) error {
@@ -15,12 +20,18 @@ func Register(eng *engine.Engine) error {
if err := remote(eng); err != nil {
return err
}
if err := eng.Register("version", dockerVersion); err != nil {
return err
}
return registry.NewService().Install(eng)
}
// remote: a RESTful api for cross-docker communication
func remote(eng *engine.Engine) error {
return eng.Register("serveapi", api.ServeApi)
if err := eng.Register("serveapi", apiserver.ServeApi); err != nil {
return err
}
return eng.Register("acceptconnections", apiserver.AcceptConnections)
}
// daemon: a default execution and storage backend for Docker on Linux,
@@ -44,3 +55,21 @@ func daemon(eng *engine.Engine) error {
}
return eng.Register("init_networkdriver", bridge.InitDriver)
}
// builtins jobs independent of any subsystem
func dockerVersion(job *engine.Job) engine.Status {
v := &engine.Env{}
v.Set("Version", dockerversion.VERSION)
v.SetJson("ApiVersion", api.APIVERSION)
v.Set("GitCommit", dockerversion.GITCOMMIT)
v.Set("GoVersion", runtime.Version())
v.Set("Os", runtime.GOOS)
v.Set("Arch", runtime.GOARCH)
if kernelVersion, err := utils.GetKernelVersion(); err == nil {
v.Set("KernelVersion", kernelVersion.String())
}
if _, err := v.WriteTo(job.Stdout); err != nil {
return job.Error(err)
}
return engine.StatusOK
}
+1 -1
View File
@@ -116,7 +116,7 @@ fi
flags=(
NAMESPACES {NET,PID,IPC,UTS}_NS
DEVPTS_MULTIPLE_INSTANCES
CGROUPS CGROUP_DEVICE
CGROUPS CGROUP_CPUACCT CGROUP_DEVICE CGROUP_SCHED
MACVLAN VETH BRIDGE
NF_NAT_IPV4 IP_NF_TARGET_MASQUERADE
NETFILTER_XT_MATCH_{ADDRTYPE,CONNTRACK}
+1 -1
View File
@@ -539,7 +539,7 @@ _docker_search()
case "$cur" in
-*)
COMPREPLY=( $( compgen -W "--no-trunc -t --trusted -s --stars" -- "$cur" ) )
COMPREPLY=( $( compgen -W "--no-trunc --automated -s --stars" -- "$cur" ) )
;;
*)
;;
+2 -2
View File
@@ -71,7 +71,7 @@ complete -c docker -A -f -n '__fish_seen_subcommand_from attach' -l sig-proxy -d
complete -c docker -A -f -n '__fish_seen_subcommand_from attach' -a '(__fish_print_docker_containers running)' -d "Container"
# build
complete -c docker -f -n '__fish_docker_no_subcommand' -a build -d 'Build a container from a Dockerfile'
complete -c docker -f -n '__fish_docker_no_subcommand' -a build -d 'Build an image from a Dockerfile'
complete -c docker -A -f -n '__fish_seen_subcommand_from build' -l no-cache -d 'Do not use cache when building the image'
complete -c docker -A -f -n '__fish_seen_subcommand_from build' -s q -l quiet -d 'Suppress the verbose output generated by the containers'
complete -c docker -A -f -n '__fish_seen_subcommand_from build' -l rm -d 'Remove intermediate containers after a successful build'
@@ -229,7 +229,7 @@ complete -c docker -A -f -n '__fish_seen_subcommand_from save' -a '(__fish_print
complete -c docker -f -n '__fish_docker_no_subcommand' -a search -d 'Search for an image in the docker index'
complete -c docker -A -f -n '__fish_seen_subcommand_from search' -l no-trunc -d "Don't truncate output"
complete -c docker -A -f -n '__fish_seen_subcommand_from search' -s s -l stars -d 'Only displays with at least xxx stars'
complete -c docker -A -f -n '__fish_seen_subcommand_from search' -s t -l trusted -d 'Only show trusted builds'
complete -c docker -A -f -n '__fish_seen_subcommand_from search' -l automated -d 'Only show automated builds'
# start
complete -c docker -f -n '__fish_docker_no_subcommand' -a start -d 'Start a stopped container'
+1 -1
View File
@@ -6,7 +6,7 @@
# /data volume is owned by sysadmin.
# USAGE:
# # Download data Dockerfile
# wget http://raw.github.com/dotcloud/docker/master/contrib/desktop-integration/data/Dockerfile
# wget http://raw.githubusercontent.com/dotcloud/docker/master/contrib/desktop-integration/data/Dockerfile
#
# # Build data image
# docker build -t data .
@@ -7,7 +7,7 @@
# sound devices. Tested on Debian 7.2
# USAGE:
# # Download Iceweasel Dockerfile
# wget http://raw.github.com/dotcloud/docker/master/contrib/desktop-integration/iceweasel/Dockerfile
# wget http://raw.githubusercontent.com/dotcloud/docker/master/contrib/desktop-integration/iceweasel/Dockerfile
#
# # Build iceweasel image
# docker build -t iceweasel .
+2
View File
@@ -4,6 +4,8 @@
# Provides: docker
# Required-Start: $syslog $remote_fs
# Required-Stop: $syslog $remote_fs
# Should-Start: cgroupfs-mount cgroup-lite
# Should-Stop: cgroupfs-mount cgroup-lite
# Default-Start: 2 3 4 5
# Default-Stop: 0 1 6
# Short-Description: Create lightweight, portable, self-sufficient containers.
+8 -1
View File
@@ -3,7 +3,7 @@
# /etc/rc.d/init.d/docker
#
# Daemon for docker.io
#
#
# chkconfig: 2345 95 95
# description: Daemon for docker.io
@@ -49,6 +49,13 @@ start() {
$exec -d $other_args &>> $logfile &
pid=$!
touch $lockfile
# wait up to 10 seconds for the pidfile to exist. see
# https://github.com/dotcloud/docker/issues/5359
tries=0
while [ ! -f $pidfile -a $tries -lt 10 ]; do
sleep 1
tries=$((tries + 1))
done
success
echo
else
+1 -1
View File
@@ -1,6 +1,6 @@
description "Docker daemon"
start on filesystem
start on (local-filesystems and net-device-up IFACE!=lo)
stop on runlevel [!2345]
limit nofile 524288 1048576
limit nproc 524288 1048576
+206
View File
@@ -0,0 +1,206 @@
% DOCKERFILE(5) Docker User Manuals
% Zac Dover
% May 2014
# NAME
Dockerfile - automate the steps of creating a Docker image
# INTRODUCTION
The **Dockerfile** is a configuration file that automates the steps of creating
a Docker image. It is similar to a Makefile. Docker reads instructions from the
**Dockerfile** to automate the steps otherwise performed manually to create an
image. To build an image, create a file called **Dockerfile**. The
**Dockerfile** describes the steps taken to assemble the image. When the
**Dockerfile** has been created, call the **docker build** command, using the
path of directory that contains **Dockerfile** as the argument.
# SYNOPSIS
INSTRUCTION arguments
For example:
FROM image
# DESCRIPTION
A Dockerfile is a file that automates the steps of creating a Docker image.
A Dockerfile is similar to a Makefile.
# USAGE
**sudo docker build .**
-- runs the steps and commits them, building a final image
The path to the source repository defines where to find the context of the
build. The build is run by the docker daemon, not the CLI. The whole
context must be transferred to the daemon. The Docker CLI reports
"Sending build context to Docker daemon" when the context is sent to the daemon.
**sudo docker build -t repository/tag .**
-- specifies a repository and tag at which to save the new image if the build
succeeds. The Docker daemon runs the steps one-by-one, commiting the result
to a new image if necessary before finally outputting the ID of the new
image. The Docker daemon automatically cleans up the context it is given.
Docker re-uses intermediate images whenever possible. This significantly
accelerates the *docker build* process.
# FORMAT
**FROM image**
or
**FROM image:tag**
-- The FROM instruction sets the base image for subsequent instructions. A
valid Dockerfile must have FROM as its first instruction. The image can be any
valid image. It is easy to start by pulling an image from the public
repositories.
-- FROM must be he first non-comment instruction in Dockerfile.
-- FROM may appear multiple times within a single Dockerfile in order to create
multiple images. Make a note of the last image id output by the commit before
each new FROM command.
-- If no tag is given to the FROM instruction, latest is assumed. If the used
tag does not exist, an error is returned.
**MAINTAINER**
--The MAINTAINER instruction sets the Author field for the generated images.
**RUN**
--RUN has two forms:
**RUN <command>**
-- (the command is run in a shell - /bin/sh -c)
**RUN ["executable", "param1", "param2"]**
--The above is executable form.
--The RUN instruction executes any commands in a new layer on top of the
current image and commits the results. The committed image is used for the next
step in Dockerfile.
--Layering RUN instructions and generating commits conforms to the core
concepts of Docker where commits are cheap and containers can be created from
any point in the history of an image. This is similar to source control. The
exec form makes it possible to avoid shell string munging. The exec form makes
it possible to RUN commands using a base image that does not contain /bin/sh.
**CMD**
--CMD has three forms:
**CMD ["executable", "param1", "param2"]** This is the preferred form, the
exec form.
**CMD ["param1", "param2"]** This command provides default parameters to
ENTRYPOINT)
**CMD command param1 param2** This command is run as a shell.
--There can be only one CMD in a Dockerfile. If more than one CMD is listed, only
the last CMD takes effect.
The main purpose of a CMD is to provide defaults for an executing container.
These defaults may include an executable, or they can omit the executable. If
they omit the executable, an ENTRYPOINT must be specified.
When used in the shell or exec formats, the CMD instruction sets the command to
be executed when running the image.
If you use the shell form of of the CMD, the <command> executes in /bin/sh -c:
**FROM ubuntu**
**CMD echo "This is a test." | wc -**
If you run <command> wihtout a shell, then you must express the command as a
JSON arry and give the full path to the executable. This array form is the
preferred form of CMD. All additional parameters must be individually expressed
as strings in the array:
**FROM ubuntu**
**CMD ["/usr/bin/wc","--help"]**
To make the container run the same executable every time, use ENTRYPOINT in
combination with CMD.
If the user specifies arguments to docker run, the specified commands override
the default in CMD.
Do not confuse **RUN** with **CMD**. RUN runs a command and commits the result. CMD
executes nothing at build time, but specifies the intended command for the
image.
**EXPOSE**
--**EXPOSE <port> [<port>...]**
The **EXPOSE** instruction informs Docker that the container listens on the
specified network ports at runtime. Docker uses this information to
interconnect containers using links, and to set up port redirection on the host
system.
**ENV**
--**ENV <key> <value>**
The ENV instruction sets the environment variable <key> to
the value <value>. This value is passed to all future RUN instructions. This is
functionally equivalent to prefixing the command with **<key>=<value>**. The
environment variables that are set with ENV persist when a container is run
from the resulting image. Use docker inspect to inspect these values, and
change them using docker run **--env <key>=<value>.**
Note that setting Setting **ENV DEBIAN_FRONTEND noninteractive** may cause
unintended consequences, because it will persist when the container is run
interactively, as with the following command: **docker run -t -i image bash**
**ADD**
--**ADD <src> <dest>** The ADD instruction copies new files from <src> and adds them
to the filesystem of the container at path <dest>. <src> must be the path to a
file or directory relative to the source directory that is being built (the
context of the build) or a remote file URL. <dest> is the absolute path to
which the source is copied inside the target container. All new files and
directories are created with mode 0755, with uid and gid 0.
**ENTRYPOINT**
--**ENTRYPOINT** has two forms: ENTRYPOINT ["executable", "param1", "param2"]
(This is like an exec, and is the preferred form.) ENTRYPOINT command param1
param2 (This is running as a shell.) An ENTRYPOINT helps you configure a
container that can be run as an executable. When you specify an ENTRYPOINT,
the whole container runs as if it was only that executable. The ENTRYPOINT
instruction adds an entry command that is not overwritten when arguments are
passed to docker run. This is different from the behavior of CMD. This allows
arguments to be passed to the entrypoint, for instance docker run <image> -d
passes the -d argument to the ENTRYPOINT. Specify parameters either in the
ENTRYPOINT JSON array (as in the preferred exec form above), or by using a CMD
statement. Parameters in the ENTRYPOINT are not overwritten by the docker run
arguments. Parameters specifies via CMD are overwritten by docker run
arguments. Specify a plain string for the ENTRYPOINT, and it will execute in
/bin/sh -c, like a CMD instruction:
FROM ubuntu
ENTRYPOINT wc -l -
This means that the Dockerfile's image always takes stdin as input (that's
what "-" means), and prints the number of lines (that's what "-l" means). To
make this optional but default, use a CMD:
FROM ubuntu
CMD ["-l", "-"]
ENTRYPOINT ["/usr/bin/wc"]
**VOLUME**
--**VOLUME ["/data"]**
The VOLUME instruction creates a mount point with the specified name and marks
it as holding externally-mounted volumes from the native host or from other
containers.
**USER**
-- **USER daemon**
The USER instruction sets the username or UID that is used when running the
image.
**WORKDIR**
-- **WORKDIR /path/to/workdir**
The WORKDIR instruction sets the working directory for the **RUN**, **CMD**, and **ENTRYPOINT** Dockerfile commands that follow it.
It can be used multiple times in a single Dockerfile. Relative paths are defined relative to the path of the previous **WORKDIR** instruction. For example:
**WORKDIR /a WORKDIR /b WORKDIR c RUN pwd**
In the above example, the output of the **pwd** command is **a/b/c**.
**ONBUILD**
-- **ONBUILD [INSTRUCTION]**
The ONBUILD instruction adds a trigger instruction to the image, which is
executed at a later time, when the image is used as the base for another
build. The trigger is executed in the context of the downstream build, as
if it had been inserted immediately after the FROM instruction in the
downstream Dockerfile. Any build instruction can be registered as a
trigger. This is useful if you are building an image to be
used as a base for building other images, for example an application build
environment or a daemon to be customized with a user-specific
configuration. For example, if your image is a reusable python
application builder, it requires application source code to be
added in a particular directory, and might require a build script
to be called after that. You can't just call ADD and RUN now, because
you don't yet have access to the application source code, and it
is different for each application build. Providing
application developers with a boilerplate Dockerfile to copy-paste
into their application is inefficient, error-prone, and
difficult to update because it mixes with application-specific code.
The solution is to use **ONBUILD** to register instructions in advance, to
run later, during the next build stage.
# HISTORY
*May 2014, Compiled by Zac Dover (zdover at redhat dot com) based on docker.io Dockerfile documentation.
+4 -3
View File
@@ -9,10 +9,11 @@ docker-attach - Attach to a running container
# DESCRIPTION
If you **docker run** a container in detached mode (**-d**), you can reattach to
the detached container with **docker attach** using the container's ID or name.
the detached container with **docker attach** using the container's ID or name.
You can detach from the container again (and leave it running) with `CTRL-c` (for
a quiet exit) or `CTRL-\` to get a stacktrace of the Docker client when it quits.
You can detach from the container again (and leave it running) with `CTRL-q
CTRL-q` (for a quiet exit), or `CTRL-c` which will send a SIGKILL to the
container, or `CTRL-\` to get a stacktrace of the Docker client when it quits.
When you detach from a container the exit code will be returned to
the client.
+39 -4
View File
@@ -2,7 +2,7 @@
% William Henry
% APRIL 2014
# NAME
docker-build - Build a container image from a Dockerfile source at PATH
docker-build - Build an image from a Dockerfile source at PATH
# SYNOPSIS
**docker build** [**--no-cache**[=*false*]] [**-q**|**--quiet**[=*false*]]
@@ -17,7 +17,7 @@ be used by **ADD** commands found within the Dockerfile.
Warning, this will send a lot of data to the Docker daemon depending
on the contents of the current directory. The build is run by the Docker
daemon, not by the CLI, so the whole context must be transferred to the daemon.
The Docker CLI reports "Uploading context" when the context is sent to
The Docker CLI reports "Sending build context to Docker daemon" when the context is sent to
the daemon.
When a single Dockerfile is given as the URL, then no context is set.
@@ -34,8 +34,9 @@ as context.
build process. The default is true.
**-t**, **--tag**=*tag*
Tag to be applied to the resulting image on successful completion of
the build.
The name to be applied to the resulting image on successful completion of
the build. `tag` in this context means the entire image name including the
optional TAG after the ':'.
**--no-cache**=*true*|*false*
When set to true, do not use a cache when building the image. The
@@ -66,6 +67,40 @@ in the Dockerfile. Note: If you include a tar file (a good practice!), then
Docker will automatically extract the contents of the tar file
specified within the `ADD` instruction into the specified target.
## Building an image and naming that image
A good practice is to give a name to the image you are building. There are
no hard rules here but it is best to give the names consideration.
The **-t**/**--tag** flag is used to rename an image. Here are some examples:
Though it is not a good practice, image names can be arbtrary:
docker build -t myimage .
A better approach is to provide a fully qualified and meaningful repository,
name, and tag (where the tag in this context means the qualifier after
the ":"). In this example we build a JBoss image for the Fedora repository
and give it the version 1.0:
docker build -t fedora/jboss:1.0
The next example is for the "whenry" user repository and uses Fedora and
JBoss and gives it the version 2.1 :
docker build -t whenry/fedora-jboss:V2.1
If you do not provide a version tag then Docker will assign `latest`:
docker build -t whenry/fedora-jboss
When you list the images, the image above will have the tag `latest`.
So renaming an image is arbitrary but consideration should be given to
a useful convention that makes sense for consumers and should also take
into account Docker community conventions.
## Building an image using a URL
This will clone the specified Github repository from the URL and use it
+25 -12
View File
@@ -14,8 +14,8 @@ docker-run - Run a process in an isolated container
[**-e**|**--env**=*environment*] [**--entrypoint**=*command*]
[**--expose**=*port*] [**-P**|**--publish-all**[=*false*]]
[**-p**|**--publish**=*port-mappping*] [**-h**|**--hostname**=*hostname*]
[**--rm**[=*false*]] [**--priviledged**[=*false*]
[**-i**|**--interactive**[=*false*]
[**--rm**[=*false*]] [**--privileged**[=*false*]]
[**-i**|**--interactive**[=*false*]]
[**-t**|**--tty**[=*false*]] [**--lxc-conf**=*options*]
[**-n**|**--networking**[=*true*]]
[**-v**|**--volume**=*volume*] [**--volumes-from**=*container-id*]
@@ -64,6 +64,9 @@ the other shell to view a list of the running containers. You can reattach to a
detached container with **docker attach**. If you choose to run a container in
the detached mode, then you cannot use the **-rm** option.
When attached in the tty mode, you can detach from a running container without
stopping the process by pressing the keys CTRL-P CTRL-Q.
**--dns**=*IP-address*
Set custom DNS servers. This option can be used to override the DNS
@@ -100,8 +103,8 @@ container can be started with the **--link**.
**-m**, **-memory**=*memory-limit*
Allows you to constrain the memory available to a container. If the host
supports swap memory, then the -m memory setting can be larger than physical
RAM. The memory limit format: <number><optional unit>, where unit = b, k, m or
g.
RAM. If a limit of 0 is specified, the container's memory is not limited. The
memory limit format: <number><optional unit>, where unit = b, k, m or g.
**-P**, **-publish-all**=*true*|*false*
When set to true publish all exposed ports to the host interfaces. The
@@ -164,7 +167,7 @@ and foreground Docker containers.
Docker container. This is because by default a container is not allowed to
access any devices. A “privileged” container is given access to all devices.
When the operator executes **docker run -privileged**, Docker will enable access
When the operator executes **docker run --privileged**, Docker will enable access
to all devices on the host as well as set some configuration in AppArmor to
allow the container nearly all the same access to the host as processes running
outside of a container on the host.
@@ -190,18 +193,28 @@ interactive shell. The default is value is false.
Set a username or UID for the container.
**-v**, **-volume**=*volume*
Bind mount a volume to the container. The **-v** option can be used one or
**-v**, **-volume**=*volume*[:ro|:rw]
Bind mount a volume to the container.
The **-v** option can be used one or
more times to add one or more mounts to a container. These mounts can then be
used in other containers using the **--volumes-from** option. See examples.
used in other containers using the **--volumes-from** option.
The volume may be optionally suffixed with :ro or :rw to mount the volumes in
read-only or read-write mode, respectively. By default, the volumes are mounted
read-write. See examples.
**--volumes-from**=*container-id*
**--volumes-from**=*container-id*[:ro|:rw]
Will mount volumes from the specified container identified by container-id.
Once a volume is mounted in a one container it can be shared with other
containers using the **--volumes-from** option when running those other
containers. The volumes can be shared even if the original container with the
mount is not running.
mount is not running.
The container ID may be optionally suffixed with :ro or
:rw to mount the volumes in read-only or read-write mode, respectively. By
default, the volumes are mounted in the same mode (read write or read only) as
the reference container.
**-w**, **-workdir**=*directory*
@@ -227,7 +240,7 @@ can override the working directory by using the **-w** option.
## Exposing log messages from the container to the host's log
If you want messages that are logged in your container to show up in the host's
syslog/journal then you should bind mount the /var/log directory as follows.
syslog/journal then you should bind mount the /dev/log directory as follows.
# docker run -v /dev/log:/dev/log -i -t fedora /bin/bash
@@ -307,7 +320,7 @@ fedora-data image:
# docker run --name=data -v /var/volume1 -v /tmp/volume2 -i -t fedora-data true
# docker run --volumes-from=data --name=fedora-container1 -i -t fedora bash
Multiple -volumes-from parameters will bring together multiple data volumes from
Multiple --volumes-from parameters will bring together multiple data volumes from
multiple containers. And it's possible to mount the volumes that came from the
DATA container in yet another container via the fedora-container1 intermidiery
container, allowing to abstract the actual data source from users of that data:
+8 -8
View File
@@ -5,7 +5,7 @@
docker-search - Search the docker index for images
# SYNOPSIS
**docker search** **--no-trunc**[=*false*] **-t**|**--trusted**[=*false*]
**docker search** **--no-trunc**[=*false*] **--automated**[=*false*]
**-s**|**--stars**[=*0*] TERM
# DESCRIPTION
@@ -13,7 +13,7 @@ docker-search - Search the docker index for images
Search an index for an image with that matches the term TERM. The table
of images returned displays the name, description (truncated by default),
number of stars awarded, whether the image is official, and whether it
is trusted.
is automated.
# OPTIONS
**--no-trunc**=*true*|*false*
@@ -23,8 +23,8 @@ is trusted.
Only displays with at least NUM (integer) stars. I.e. only those images
ranked >=NUM.
**-t**, **--trusted**=*true*|*false*
When true only show trusted builds. The default is false.
**--automated**=*true*|*false*
When true only show automated builds. The default is false.
# EXAMPLE
@@ -34,19 +34,19 @@ Search the registry for the term 'fedora' and only display those images
ranked 3 or higher:
$ sudo docker search -s 3 fedora
NAME DESCRIPTION STARS OFFICIAL TRUSTED
NAME DESCRIPTION STARS OFFICIAL AUTOMATED
mattdm/fedora A basic Fedora image corresponding roughly... 50
fedora (Semi) Official Fedora base image. 38
mattdm/fedora-small A small Fedora image on which to build. Co... 8
goldmann/wildfly A WildFly application server running on a ... 3 [OK]
## Search the registry for trusted images
## Search the registry for automated images
Search the registry for the term 'fedora' and only display trusted images
Search the registry for the term 'fedora' and only display automated images
ranked 1 or higher:
$ sudo docker search -s 1 -t fedora
NAME DESCRIPTION STARS OFFICIAL TRUSTED
NAME DESCRIPTION STARS OFFICIAL AUTOMATED
goldmann/wildfly A WildFly application server running on a ... 3 [OK]
tutum/fedora-20 Fedora 20 image with SSH access. For the r... 1 [OK]
+4
View File
@@ -20,6 +20,10 @@ the process
**-i**, **--interactive**=*true*|*false*
When true attach to container's stdin
# NOTES
If run on a started container, start takes no action and succeeds
unconditionally.
# HISTORY
April 2014, Originally compiled by William Henry (whenry at redhat dot com)
based on docker.io source material and internal work.
+9 -5
View File
@@ -9,11 +9,12 @@ docker-tag - Tag an image in the repository
IMAGE [REGISTRYHOST/][USERNAME/]NAME[:TAG]
# DESCRIPTION
This will tag an image in the repository.
This will give a new alias to an image in the repository. This refers to the
entire image name including the optional TAG after the ':'.
# "OPTIONS"
**-f**, **--force**=*true*|*false*
When set to true, force the tag name. The default is *false*.
When set to true, force the alias. The default is *false*.
**REGISTRYHOST**
The hostname of the registry if required. This may also include the port
@@ -26,13 +27,16 @@ separated by a ':'
The image name.
**TAG**
The tag you are assigning to the image.
The tag you are assigning to the image. Though this is arbitrary it is
recommended to be used for a version to disinguish images with the same name.
Note that here TAG is a part of the overall name or "tag".
# EXAMPLES
## Tagging an image
## Giving an image a new alias
Here is an example of tagging an image with the tag version1.0 :
Here is an example of aliasing an image (e.g. 0e5574283393) as "httpd" and
tagging it into the "fedora" repository with "version1.0":
docker tag 0e5574283393 fedora/httpd:version1.0
+5 -5
View File
@@ -26,10 +26,10 @@ To see the man page for a command run **man docker <command>**.
**-D**=*true*|*false*
Enable debug mode. Default is false.
**-H**, **--host**=[unix:///var/run/docker.sock]: tcp://[host[:port]] to bind or
**-H**, **--host**=[unix:///var/run/docker.sock]: tcp://[host:port] to bind or
unix://[/path/to/socket] to use.
Enable both the socket support and TCP on localhost. When host=[0.0.0.0],
port=[4243] or path =[/var/run/docker.sock] is omitted, default values are used.
The socket(s) to bind to in daemon mode specified using one or more
tcp://host:port, unix:///path/to/socket, fd://* or fd://socketfd.
**--api-enable-cors**=*true*|*false*
Enable CORS headers in the remote API. Default is false.
@@ -73,7 +73,7 @@ port=[4243] or path =[/var/run/docker.sock] is omitted, default values are used.
**-v**=*true*|*false*
Print version information and quit. Default is false.
**--selinux-enabled=*true*|*false*
**--selinux-enabled**=*true*|*false*
Enable selinux support. Default is false.
# COMMANDS
@@ -81,7 +81,7 @@ port=[4243] or path =[/var/run/docker.sock] is omitted, default values are used.
Attach to a running container
**docker-build(1)**
Build a container from a Dockerfile
Build an image from a Dockerfile
**docker-commit(1)**
Create a new image from a container's changes
+1 -1
View File
@@ -3,7 +3,7 @@
.\"
.TH "DOCKER" "1" "MARCH 2014" "0.1" "Docker"
.SH NAME
docker-build \- Build a container image from a Dockerfile source at PATH
docker-build \- Build an image from a Dockerfile source at PATH
.SH SYNOPSIS
.B docker build
[\fB--no-cache\fR[=\fIfalse\fR]
+2 -2
View File
@@ -39,7 +39,7 @@ CPU shares in relative weight. You can increase the priority of a container with
.TP
.B -m, --memory=\fImemory-limit\fR:
Allows you to constrain the memory available to a container. If the host supports swap memory, then the -m memory setting can be larger than physical RAM. The memory limit format: <number><optional unit>, where unit = b, k, m or g.
Allows you to constrain the memory available to a container. If the host supports swap memory, then the -m memory setting can be larger than physical RAM. If a limit of 0 is specified, the container's memory is not limited. The memory limit format: <number><optional unit>, where unit = b, k, m or g.
.TP
.B --cidfile=\fIfile\fR:
@@ -245,7 +245,7 @@ docker run --volumes-from=data --name=fedora-container1 -i -t fedora bash
.RE
.sp
.TP
Multiple -volumes-from parameters will bring together multiple data volumes from multiple containers. And it's possible to mount the volumes that came from the DATA container in yet another container via the fedora-container1 intermidiery container, allowing to abstract the actual data source from users of that data:
Multiple --volumes-from parameters will bring together multiple data volumes from multiple containers. And it's possible to mount the volumes that came from the DATA container in yet another container via the fedora-container1 intermidiery container, allowing to abstract the actual data source from users of that data:
.sp
.RS
docker run --volumes-from=fedora-container1 --name=fedora-container2 -i -t fedora bash
+2 -2
View File
@@ -19,7 +19,7 @@ To see the man page for a command run \fBman docker <command>\fR.
Enable debug mode
.TP
.B\-H=[unix:///var/run/docker.sock]: tcp://[host[:port]] to bind or unix://[/path/to/socket] to use.
When host=[0.0.0.0], port=[4243] or path
When host=[0.0.0.0], port=[2375] or path
=[/var/run/docker.sock] is omitted, default values are used.
.TP
.B \-\-api-enable-cors=false
@@ -69,7 +69,7 @@ Print version information and quit
Attach to a running container
.TP
.B build
Build a container from a Dockerfile
Build an image from a Dockerfile
.TP
.B commit
Create a new image from a container's changes
+4
View File
@@ -2,6 +2,10 @@
# Generate a very minimal filesystem based on busybox-static,
# and load it into the local docker under the name "busybox".
echo >&2
echo >&2 'warning: this script is deprecated - see mkimage.sh and mkimage/busybox-static'
echo >&2
BUSYBOX=$(which busybox)
[ "$BUSYBOX" ] || {
echo "Sorry, I could not locate busybox."
+4
View File
@@ -1,6 +1,10 @@
#!/usr/bin/env bash
set -e
echo >&2
echo >&2 'warning: this script is deprecated - see mkimage.sh and mkimage/debootstrap'
echo >&2
variant='minbase'
include='iproute,iputils-ping'
arch='amd64' # intentionally undocumented for now
+4
View File
@@ -8,6 +8,10 @@
set -e
echo >&2
echo >&2 'warning: this script is deprecated - see mkimage.sh and mkimage/rinse'
echo >&2
repo="$1"
distro="$2"
mirror="$3"
+105
View File
@@ -0,0 +1,105 @@
#!/usr/bin/env bash
set -e
mkimg="$(basename "$0")"
usage() {
echo >&2 "usage: $mkimg [-d dir] [-t tag] script [script-args]"
echo >&2 " ie: $mkimg -t someuser/debian debootstrap --variant=minbase jessie"
echo >&2 " $mkimg -t someuser/ubuntu debootstrap --include=ubuntu-minimal trusty"
echo >&2 " $mkimg -t someuser/busybox busybox-static"
echo >&2 " $mkimg -t someuser/centos:5 rinse --distribution centos-5"
exit 1
}
scriptDir="$(dirname "$(readlink -f "$BASH_SOURCE")")/mkimage"
optTemp=$(getopt --options '+d:t:h' --longoptions 'dir:,tag:,help' --name "$mkimg" -- "$@")
eval set -- "$optTemp"
unset optTemp
dir=
tag=
while true; do
case "$1" in
-d|--dir) dir="$2" ; shift 2 ;;
-t|--tag) tag="$2" ; shift 2 ;;
-h|--help) usage ;;
--) shift ; break ;;
esac
done
script="$1"
[ "$script" ] || usage
shift
if [ ! -x "$scriptDir/$script" ]; then
echo >&2 "error: $script does not exist or is not executable"
echo >&2 " see $scriptDir for possible scripts"
exit 1
fi
# don't mistake common scripts like .febootstrap-minimize as image-creators
if [[ "$script" == .* ]]; then
echo >&2 "error: $script is a script helper, not a script"
echo >&2 " see $scriptDir for possible scripts"
exit 1
fi
delDir=
if [ -z "$dir" ]; then
dir="$(mktemp -d ${TMPDIR:-/tmp}/docker-mkimage.XXXXXXXXXX)"
delDir=1
fi
rootfsDir="$dir/rootfs"
( set -x; mkdir -p "$rootfsDir" )
# pass all remaining arguments to $script
"$scriptDir/$script" "$rootfsDir" "$@"
# Docker mounts tmpfs at /dev and procfs at /proc so we can remove them
rm -rf "$rootfsDir/dev" "$rootfsDir/proc"
mkdir -p "$rootfsDir/dev" "$rootfsDir/proc"
# make sure /etc/resolv.conf has something useful in it
mkdir -p "$rootfsDir/etc"
cat > "$rootfsDir/etc/resolv.conf" <<'EOF'
nameserver 8.8.8.8
nameserver 8.8.4.4
EOF
tarFile="$dir/rootfs.tar.xz"
touch "$tarFile"
(
set -x
tar --numeric-owner -caf "$tarFile" -C "$rootfsDir" --transform='s,^./,,' .
)
echo >&2 "+ cat > '$dir/Dockerfile'"
cat > "$dir/Dockerfile" <<'EOF'
FROM scratch
ADD rootfs.tar.xz /
EOF
# if our generated image has a decent shell, let's set a default command
for shell in /bin/bash /usr/bin/fish /usr/bin/zsh /bin/sh; do
if [ -x "$rootfsDir/$shell" ]; then
( set -x; echo 'CMD ["'"$shell"'"]' >> "$dir/Dockerfile" )
break
fi
done
( set -x; rm -rf "$rootfsDir" )
if [ "$tag" ]; then
( set -x; docker build -t "$tag" "$dir" )
elif [ "$delDir" ]; then
# if we didn't specify a tag and we're going to delete our dir, let's just build an untagged image so that we did _something_
( set -x; docker build "$dir" )
fi
if [ "$delDir" ]; then
( set -x; rm -rf "$dir" )
fi
+28
View File
@@ -0,0 +1,28 @@
#!/usr/bin/env bash
set -e
rootfsDir="$1"
shift
(
cd "$rootfsDir"
# effectively: febootstrap-minimize --keep-zoneinfo --keep-rpmdb --keep-services "$target"
# locales
rm -rf usr/{{lib,share}/locale,{lib,lib64}/gconv,bin/localedef,sbin/build-locale-archive}
# docs
rm -rf usr/share/{man,doc,info,gnome/help}
# cracklib
#rm -rf usr/share/cracklib
# i18n
rm -rf usr/share/i18n
# yum cache
rm -rf var/cache/yum
mkdir -p --mode=0755 var/cache/yum
# sln
rm -rf sbin/sln
# ldconfig
#rm -rf sbin/ldconfig
rm -rf etc/ld.so.cache var/cache/ldconfig
mkdir -p --mode=0755 var/cache/ldconfig
)
+34
View File
@@ -0,0 +1,34 @@
#!/usr/bin/env bash
set -e
rootfsDir="$1"
shift
busybox="$(which busybox 2>/dev/null || true)"
if [ -z "$busybox" ]; then
echo >&2 'error: busybox: not found'
echo >&2 ' install it with your distribution "busybox-static" package'
exit 1
fi
if ! ldd "$busybox" 2>&1 | grep -q 'not a dynamic executable'; then
echo >&2 "error: '$busybox' appears to be a dynamic executable"
echo >&2 ' you should install your distribution "busybox-static" package instead'
exit 1
fi
mkdir -p "$rootfsDir/bin"
rm -f "$rootfsDir/bin/busybox" # just in case
cp "$busybox" "$rootfsDir/bin/busybox"
(
cd "$rootfsDir"
IFS=$'\n'
modules=( $(bin/busybox --list-modules) )
unset IFS
for module in "${modules[@]}"; do
mkdir -p "$(dirname "$module")"
ln -sf /bin/busybox "$module"
done
)
+125
View File
@@ -0,0 +1,125 @@
#!/usr/bin/env bash
set -e
rootfsDir="$1"
shift
# we have to do a little fancy footwork to make sure "rootfsDir" becomes the second non-option argument to debootstrap
before=()
while [ $# -gt 0 ] && [[ "$1" == -* ]]; do
before+=( "$1" )
shift
done
suite="$1"
shift
(
set -x
debootstrap "${before[@]}" "$suite" "$rootfsDir" "$@"
)
# now for some Docker-specific tweaks
# prevent init scripts from running during install/update
echo >&2 "+ cat > '$rootfsDir/usr/sbin/policy-rc.d'"
cat > "$rootfsDir/usr/sbin/policy-rc.d" <<'EOF'
#!/bin/sh
exit 101
EOF
chmod +x "$rootfsDir/usr/sbin/policy-rc.d"
# prevent upstart scripts from running during install/update
(
set -x
chroot "$rootfsDir" dpkg-divert --local --rename --add /sbin/initctl
ln -sf /bin/true "$rootfsDir/sbin/initctl"
)
# shrink the image, since apt makes us fat (wheezy: ~157.5MB vs ~120MB)
( set -x; chroot "$rootfsDir" apt-get clean )
# Ubuntu 10.04 sucks... :)
if strings "$rootfsDir/usr/bin/dpkg" | grep -q unsafe-io; then
# force dpkg not to call sync() after package extraction (speeding up installs)
echo >&2 "+ echo force-unsafe-io > '$rootfsDir/etc/dpkg/dpkg.cfg.d/docker-apt-speedup'"
echo 'force-unsafe-io' > "$rootfsDir/etc/dpkg/dpkg.cfg.d/docker-apt-speedup"
fi
if [ -d "$rootfsDir/etc/apt/apt.conf.d" ]; then
# _keep_ us lean by effectively running "apt-get clean" after every install
aptGetClean='"rm -f /var/cache/apt/archives/*.deb /var/cache/apt/archives/partial/*.deb /var/cache/apt/*.bin || true";'
echo >&2 "+ cat > '$rootfsDir/etc/apt/apt.conf.d/docker-clean'"
cat > "$rootfsDir/etc/apt/apt.conf.d/docker-clean" <<-EOF
DPkg::Post-Invoke { ${aptGetClean} };
APT::Update::Post-Invoke { ${aptGetClean} };
Dir::Cache::pkgcache "";
Dir::Cache::srcpkgcache "";
EOF
# remove apt-cache translations for fast "apt-get update"
echo >&2 "+ cat > '$rootfsDir/etc/apt/apt.conf.d/docker-no-languages'"
echo 'Acquire::Languages "none";' > "$rootfsDir/etc/apt/apt.conf.d/docker-no-languages"
fi
if [ -z "$DONT_TOUCH_SOURCES_LIST" ]; then
# tweak sources.list, where appropriate
lsbDist=
if [ -z "$lsbDist" -a -r "$rootfsDir/etc/os-release" ]; then
lsbDist="$(. "$rootfsDir/etc/os-release" && echo "$ID")"
fi
if [ -z "$lsbDist" -a -r "$rootfsDir/etc/lsb-release" ]; then
lsbDist="$(. "$rootfsDir/etc/lsb-release" && echo "$DISTRIB_ID")"
fi
if [ -z "$lsbDist" -a -r "$rootfsDir/etc/debian_version" ]; then
lsbDist='Debian'
fi
case "$lsbDist" in
debian|Debian)
# updates and security!
if [ "$suite" != 'sid' -a "$suite" != 'unstable' ]; then
(
set -x
sed -i "p; s/ $suite main$/ ${suite}-updates main/" "$rootfsDir/etc/apt/sources.list"
echo "deb http://security.debian.org $suite/updates main" >> "$rootfsDir/etc/apt/sources.list"
)
fi
;;
ubuntu|Ubuntu)
# add the universe, updates, and security repositories
(
set -x
sed -i "
s/ $suite main$/ $suite main universe/; p;
s/ $suite main/ ${suite}-updates main/; p;
s/ $suite-updates main/ ${suite}-security main/
" "$rootfsDir/etc/apt/sources.list"
)
;;
tanglu|Tanglu)
# add the updates repository
if [ "$suite" != 'devel' ]; then
(
set -x
sed -i "p; s/ $suite main$/ ${suite}-updates main/" "$rootfsDir/etc/apt/sources.list"
)
fi
;;
steamos|SteamOS)
# add contrib and non-free
(
set -x
sed -i "s/ $suite main$/ $suite main contrib non-free/" "$rootfsDir/etc/apt/sources.list"
)
;;
esac
fi
# make sure we're fully up-to-date, too
(
set -x
chroot "$rootfsDir" apt-get update
chroot "$rootfsDir" apt-get dist-upgrade -y
)
+25
View File
@@ -0,0 +1,25 @@
#!/usr/bin/env bash
set -e
rootfsDir="$1"
shift
# specifying --arch below is safe because "$@" can override it and the "latest" one wins :)
(
set -x
rinse --directory "$rootfsDir" --arch amd64 "$@"
)
"$(dirname "$BASH_SOURCE")/.febootstrap-minimize" "$rootfsDir"
if [ -d "$rootfsDir/etc/sysconfig" ]; then
# allow networking init scripts inside the container to work without extra steps
echo 'NETWORKING=yes' > "$rootfsDir/etc/sysconfig/network"
fi
# make sure we're fully up-to-date, too
(
set -x
chroot "$rootfsDir" yum update -y
)
+4 -4
View File
@@ -31,20 +31,20 @@ stop on runlevel [!2345]
respawn
script
/usr/bin/docker -d -H=tcp://0.0.0.0:4243
/usr/bin/docker -d -H=tcp://0.0.0.0:2375
end script
```
Once that's done, you need to set up a SSH tunnel between your host machine and the vagrant machine that's running Docker. This can be done by running the following command in a host terminal:
```
ssh -L 4243:localhost:4243 -p 2222 vagrant@localhost
ssh -L 2375:localhost:2375 -p 2222 vagrant@localhost
```
(The first 4243 is what your host can connect to, the second 4243 is what port Docker is running on in the vagrant machine, and the 2222 is the port Vagrant is providing for SSH. If VirtualBox is the VM you're using, you can see what value "2222" should be by going to: Network > Adapter 1 > Advanced > Port Forwarding in the VirtualBox GUI.)
(The first 2375 is what your host can connect to, the second 2375 is what port Docker is running on in the vagrant machine, and the 2222 is the port Vagrant is providing for SSH. If VirtualBox is the VM you're using, you can see what value "2222" should be by going to: Network > Adapter 1 > Advanced > Port Forwarding in the VirtualBox GUI.)
Note that because the port has been changed, to run docker commands from within the command line you must run them like this:
```
sudo docker -H 0.0.0.0:4243 < commands for docker >
sudo docker -H 0.0.0.0:2375 < commands for docker >
```
+10
View File
@@ -0,0 +1,10 @@
This directory contains code pertaining to running containers and storing images
Code pertaining to running containers:
- execdriver
- networkdriver
Code pertaining to storing images:
- graphdriver
+99 -74
View File
@@ -9,6 +9,7 @@ import (
"log"
"os"
"path"
"path/filepath"
"strings"
"sync"
"syscall"
@@ -22,8 +23,10 @@ import (
"github.com/dotcloud/docker/links"
"github.com/dotcloud/docker/nat"
"github.com/dotcloud/docker/pkg/label"
"github.com/dotcloud/docker/pkg/libcontainer/devices"
"github.com/dotcloud/docker/pkg/networkfs/etchosts"
"github.com/dotcloud/docker/pkg/networkfs/resolvconf"
"github.com/dotcloud/docker/pkg/symlink"
"github.com/dotcloud/docker/runconfig"
"github.com/dotcloud/docker/utils"
)
@@ -81,42 +84,6 @@ type Container struct {
activeLinks map[string]*links.Link
}
// Inject the io.Reader at the given path. Note: do not close the reader
func (container *Container) Inject(file io.Reader, pth string) error {
if err := container.Mount(); err != nil {
return fmt.Errorf("inject: error mounting container %s: %s", container.ID, err)
}
defer container.Unmount()
// Return error if path exists
destPath := path.Join(container.basefs, pth)
if _, err := os.Stat(destPath); err == nil {
// Since err is nil, the path could be stat'd and it exists
return fmt.Errorf("%s exists", pth)
} else if !os.IsNotExist(err) {
// Expect err might be that the file doesn't exist, so
// if it's some other error, return that.
return err
}
// Make sure the directory exists
if err := os.MkdirAll(path.Join(container.basefs, path.Dir(pth)), 0755); err != nil {
return err
}
dest, err := os.Create(destPath)
if err != nil {
return err
}
defer dest.Close()
if _, err := io.Copy(dest, file); err != nil {
return err
}
return nil
}
func (container *Container) FromDisk() error {
data, err := ioutil.ReadFile(container.jsonPath())
if err != nil {
@@ -170,6 +137,16 @@ func (container *Container) WriteHostConfig() (err error) {
return ioutil.WriteFile(container.hostConfigPath(), data, 0666)
}
func (container *Container) getResourcePath(path string) string {
cleanPath := filepath.Join("/", path)
return filepath.Join(container.basefs, cleanPath)
}
func (container *Container) getRootResourcePath(path string) string {
cleanPath := filepath.Join("/", path)
return filepath.Join(container.root, cleanPath)
}
func populateCommand(c *Container, env []string) error {
var (
en *execdriver.Network
@@ -215,20 +192,23 @@ func populateCommand(c *Container, env []string) error {
Memory: c.Config.Memory,
MemorySwap: c.Config.MemorySwap,
CpuShares: c.Config.CpuShares,
Cpuset: c.Config.Cpuset,
}
c.command = &execdriver.Command{
ID: c.ID,
Privileged: c.hostConfig.Privileged,
Rootfs: c.RootfsPath(),
InitPath: "/.dockerinit",
Entrypoint: c.Path,
Arguments: c.Args,
WorkingDir: c.Config.WorkingDir,
Network: en,
Tty: c.Config.Tty,
User: c.Config.User,
Config: context,
Resources: resources,
ID: c.ID,
Privileged: c.hostConfig.Privileged,
Rootfs: c.RootfsPath(),
InitPath: "/.dockerinit",
Entrypoint: c.Path,
Arguments: c.Args,
WorkingDir: c.Config.WorkingDir,
Network: en,
Tty: c.Config.Tty,
User: c.Config.User,
Config: context,
Resources: resources,
AllowedDevices: devices.DefaultAllowedDevices,
AutoCreatedDevices: devices.DefaultAutoCreatedDevices,
}
c.command.SysProcAttr = &syscall.SysProcAttr{Setsid: true}
c.command.Env = env
@@ -344,7 +324,7 @@ func (container *Container) StderrLogPipe() io.ReadCloser {
}
func (container *Container) buildHostnameFile() error {
container.HostnamePath = path.Join(container.root, "hostname")
container.HostnamePath = container.getRootResourcePath("hostname")
if container.Config.Domainname != "" {
return ioutil.WriteFile(container.HostnamePath, []byte(fmt.Sprintf("%s.%s\n", container.Config.Hostname, container.Config.Domainname)), 0644)
}
@@ -356,7 +336,7 @@ func (container *Container) buildHostnameAndHostsFiles(IP string) error {
return err
}
container.HostsPath = path.Join(container.root, "hosts")
container.HostsPath = container.getRootResourcePath("hosts")
extraContent := make(map[string]string)
@@ -455,6 +435,20 @@ func (container *Container) monitor(callback execdriver.StartCallback) error {
utils.Errorf("Error running container: %s", err)
}
// Cleanup
container.cleanup()
// Re-create a brand new stdin pipe once the container exited
if container.Config.OpenStdin {
container.stdin, container.stdinPipe = io.Pipe()
}
if container.daemon != nil && container.daemon.srv != nil {
container.daemon.srv.LogEvent("die", container.ID, container.daemon.repositories.ImageName(container.Image))
}
close(container.waitLock)
if container.daemon != nil && container.daemon.srv != nil && container.daemon.srv.IsRunning() {
container.State.SetStopped(exitCode)
@@ -470,20 +464,6 @@ func (container *Container) monitor(callback execdriver.StartCallback) error {
}
}
// Cleanup
container.cleanup()
// Re-create a brand new stdin pipe once the container exited
if container.Config.OpenStdin {
container.stdin, container.stdinPipe = io.Pipe()
}
if container.daemon != nil && container.daemon.srv != nil {
container.daemon.srv.LogEvent("die", container.ID, container.daemon.repositories.ImageName(container.Image))
}
close(container.waitLock)
return err
}
@@ -522,12 +502,37 @@ func (container *Container) KillSig(sig int) error {
container.Lock()
defer container.Unlock()
// We could unpause the container for them rather than returning this error
if container.State.IsPaused() {
return fmt.Errorf("Container %s is paused. Unpause the container before stopping", container.ID)
}
if !container.State.IsRunning() {
return nil
}
return container.daemon.Kill(container, sig)
}
func (container *Container) Pause() error {
if container.State.IsPaused() {
return fmt.Errorf("Container %s is already paused", container.ID)
}
if !container.State.IsRunning() {
return fmt.Errorf("Container %s is not running", container.ID)
}
return container.daemon.Pause(container)
}
func (container *Container) Unpause() error {
if !container.State.IsPaused() {
return fmt.Errorf("Container %s is not paused", container.ID)
}
if !container.State.IsRunning() {
return fmt.Errorf("Container %s is not running", container.ID)
}
return container.daemon.Unpause(container)
}
func (container *Container) Kill() error {
if !container.State.IsRunning() {
return nil
@@ -571,6 +576,7 @@ func (container *Container) Stop(seconds int) error {
log.Printf("Container %v failed to exit within %d seconds of SIGTERM - using the force", container.ID, seconds)
// 3. If it doesn't, then send SIGKILL
if err := container.Kill(); err != nil {
container.Wait()
return err
}
}
@@ -640,7 +646,7 @@ func (container *Container) Export() (archive.Archive, error) {
}
func (container *Container) WaitTimeout(timeout time.Duration) error {
done := make(chan bool)
done := make(chan bool, 1)
go func() {
container.Wait()
done <- true
@@ -659,6 +665,8 @@ func (container *Container) Mount() error {
}
func (container *Container) Changes() ([]archive.Change, error) {
container.Lock()
defer container.Unlock()
return container.daemon.Changes(container)
}
@@ -674,7 +682,7 @@ func (container *Container) Unmount() error {
}
func (container *Container) logPath(name string) string {
return path.Join(container.root, fmt.Sprintf("%s-%s.log", container.ID, name))
return container.getRootResourcePath(fmt.Sprintf("%s-%s.log", container.ID, name))
}
func (container *Container) ReadLog(name string) (io.Reader, error) {
@@ -682,11 +690,11 @@ func (container *Container) ReadLog(name string) (io.Reader, error) {
}
func (container *Container) hostConfigPath() string {
return path.Join(container.root, "hostconfig.json")
return container.getRootResourcePath("hostconfig.json")
}
func (container *Container) jsonPath() string {
return path.Join(container.root, "config.json")
return container.getRootResourcePath("config.json")
}
// This method must be exported to be used from the lxc template
@@ -745,8 +753,16 @@ func (container *Container) Copy(resource string) (io.ReadCloser, error) {
if err := container.Mount(); err != nil {
return nil, err
}
var filter []string
basePath := path.Join(container.basefs, resource)
resPath := container.getResourcePath(resource)
basePath, err := symlink.FollowSymlinkInScope(resPath, container.basefs)
if err != nil {
container.Unmount()
return nil, err
}
stat, err := os.Stat(basePath)
if err != nil {
container.Unmount()
@@ -766,6 +782,7 @@ func (container *Container) Copy(resource string) (io.ReadCloser, error) {
Includes: filter,
})
if err != nil {
container.Unmount()
return nil, err
}
return utils.NewReadCloserWrapper(archive, func() error {
@@ -844,7 +861,7 @@ func (container *Container) setupContainerDns() error {
} else if len(daemon.config.DnsSearch) > 0 {
dnsSearch = daemon.config.DnsSearch
}
container.ResolvConfPath = path.Join(container.root, "resolv.conf")
container.ResolvConfPath = container.getRootResourcePath("resolv.conf")
return resolvconf.Build(container.ResolvConfPath, dns, dnsSearch)
} else {
container.ResolvConfPath = "/etc/resolv.conf"
@@ -865,9 +882,17 @@ func (container *Container) initializeNetworking() error {
container.Config.Hostname = parts[0]
container.Config.Domainname = parts[1]
}
container.HostsPath = "/etc/hosts"
return container.buildHostnameFile()
content, err := ioutil.ReadFile("/etc/hosts")
if os.IsNotExist(err) {
return container.buildHostnameAndHostsFiles("")
}
if err != nil {
return err
}
container.HostsPath = container.getRootResourcePath("hosts")
return ioutil.WriteFile(container.HostsPath, content, 0644)
} else if container.hostConfig.NetworkMode.IsContainer() {
// we need to get the hosts files from the container to join
nc, err := container.getNetworkedContainer()
@@ -982,12 +1007,12 @@ func (container *Container) setupWorkingDirectory() error {
if container.Config.WorkingDir != "" {
container.Config.WorkingDir = path.Clean(container.Config.WorkingDir)
pthInfo, err := os.Stat(path.Join(container.basefs, container.Config.WorkingDir))
pthInfo, err := os.Stat(container.getResourcePath(container.Config.WorkingDir))
if err != nil {
if !os.IsNotExist(err) {
return err
}
if err := os.MkdirAll(path.Join(container.basefs, container.Config.WorkingDir), 0755); err != nil {
if err := os.MkdirAll(container.getResourcePath(container.Config.WorkingDir), 0755); err != nil {
return err
}
}
+198 -95
View File
@@ -1,7 +1,6 @@
package daemon
import (
"container/list"
"fmt"
"io"
"io/ioutil"
@@ -28,7 +27,7 @@ import (
"github.com/dotcloud/docker/image"
"github.com/dotcloud/docker/pkg/graphdb"
"github.com/dotcloud/docker/pkg/label"
"github.com/dotcloud/docker/pkg/mount"
"github.com/dotcloud/docker/pkg/namesgenerator"
"github.com/dotcloud/docker/pkg/networkfs/resolvconf"
"github.com/dotcloud/docker/pkg/selinux"
"github.com/dotcloud/docker/pkg/sysinfo"
@@ -47,10 +46,43 @@ var (
validContainerNamePattern = regexp.MustCompile(`^/?` + validContainerNameChars + `+$`)
)
type contStore struct {
s map[string]*Container
sync.Mutex
}
func (c *contStore) Add(id string, cont *Container) {
c.Lock()
c.s[id] = cont
c.Unlock()
}
func (c *contStore) Get(id string) *Container {
c.Lock()
res := c.s[id]
c.Unlock()
return res
}
func (c *contStore) Delete(id string) {
c.Lock()
delete(c.s, id)
c.Unlock()
}
func (c *contStore) List() []*Container {
containers := new(History)
for _, cont := range c.s {
containers.Add(cont)
}
containers.Sort()
return *containers
}
type Daemon struct {
repository string
sysInitPath string
containers *list.List
containers *contStore
graph *graph.Graph
repositories *graph.TagStore
idIndex *utils.TruncIndex
@@ -64,38 +96,14 @@ type Daemon struct {
execDriver execdriver.Driver
}
// Mountpoints should be private to the container
func remountPrivate(mountPoint string) error {
mounted, err := mount.Mounted(mountPoint)
if err != nil {
return err
}
if !mounted {
if err := mount.Mount(mountPoint, mountPoint, "none", "bind,rw"); err != nil {
return err
}
}
return mount.ForceMount("", mountPoint, "none", "private")
// Install installs daemon capabilities to eng.
func (daemon *Daemon) Install(eng *engine.Engine) error {
return eng.Register("container_inspect", daemon.ContainerInspect)
}
// List returns an array of all containers registered in the daemon.
func (daemon *Daemon) List() []*Container {
containers := new(History)
for e := daemon.containers.Front(); e != nil; e = e.Next() {
containers.Add(e.Value.(*Container))
}
return *containers
}
func (daemon *Daemon) getContainerElement(id string) *list.Element {
for e := daemon.containers.Front(); e != nil; e = e.Next() {
container := e.Value.(*Container)
if container.ID == id {
return e
}
}
return nil
return daemon.containers.List()
}
// Get looks for a container by the specified ID or name, and returns it.
@@ -110,11 +118,7 @@ func (daemon *Daemon) Get(name string) *Container {
return nil
}
e := daemon.getContainerElement(id)
if e == nil {
return nil
}
return e.Value.(*Container)
return daemon.containers.Get(id)
}
// Exists returns a true if a container of the specified ID or name exists,
@@ -141,7 +145,13 @@ func (daemon *Daemon) load(id string) (*Container, error) {
}
// Register makes a container object usable by the daemon as <container.ID>
// This is a wrapper for register
func (daemon *Daemon) Register(container *Container) error {
return daemon.register(container, true, nil)
}
// register makes a container object usable by the daemon as <container.ID>
func (daemon *Daemon) register(container *Container, updateSuffixarray bool, containersToStart *[]*Container) error {
if container.daemon != nil || daemon.Exists(container.ID) {
return fmt.Errorf("Container is already loaded")
}
@@ -164,8 +174,15 @@ func (daemon *Daemon) Register(container *Container) error {
container.stdinPipe = utils.NopWriteCloser(ioutil.Discard) // Silently drop stdin
}
// done
daemon.containers.PushBack(container)
daemon.idIndex.Add(container.ID)
daemon.containers.Add(container.ID, container)
// don't update the Suffixarray if we're starting up
// we'll waste time if we update it for every container
if updateSuffixarray {
daemon.idIndex.Add(container.ID)
} else {
daemon.idIndex.AddWithoutSuffixarrayUpdate(container.ID)
}
// FIXME: if the container is supposed to be running but is not, auto restart it?
// if so, then we need to restart monitor and init a new lock
@@ -203,13 +220,13 @@ func (daemon *Daemon) Register(container *Container) error {
if !info.IsRunning() {
utils.Debugf("Container %s was supposed to be running but is not.", container.ID)
if daemon.config.AutoRestart {
utils.Debugf("Restarting")
utils.Debugf("Marking as restarting")
if err := container.Unmount(); err != nil {
utils.Debugf("restart unmount error %s", err)
}
if err := container.Start(); err != nil {
return err
if containersToStart != nil {
*containersToStart = append(*containersToStart, container)
}
} else {
utils.Debugf("Marking as stopped")
@@ -231,20 +248,15 @@ func (daemon *Daemon) Register(container *Container) error {
func (daemon *Daemon) ensureName(container *Container) error {
if container.Name == "" {
name, err := generateRandomName(daemon)
name, err := daemon.generateNewName(container.ID)
if err != nil {
name = utils.TruncateID(container.ID)
return err
}
container.Name = name
if err := container.ToDisk(); err != nil {
utils.Debugf("Error saving container name %s", err)
}
if !daemon.containerGraph.Exists(name) {
if _, err := daemon.containerGraph.Set(name, container.ID); err != nil {
utils.Debugf("Setting default id - %s", err)
}
}
}
return nil
}
@@ -264,7 +276,7 @@ func (daemon *Daemon) Destroy(container *Container) error {
return fmt.Errorf("The given container is <nil>")
}
element := daemon.getContainerElement(container.ID)
element := daemon.containers.Get(container.ID)
if element == nil {
return fmt.Errorf("Container %v not found - maybe it was already destroyed?", container.ID)
}
@@ -275,7 +287,11 @@ func (daemon *Daemon) Destroy(container *Container) error {
// Deregister the container before removing its directory, to avoid race conditions
daemon.idIndex.Delete(container.ID)
daemon.containers.Remove(element)
daemon.containers.Delete(container.ID)
if _, err := daemon.containerGraph.Purge(container.ID); err != nil {
utils.Debugf("Unable to remove container from link graph: %s", err)
}
if err := daemon.driver.Remove(container.ID); err != nil {
return fmt.Errorf("Driver %s failed to remove root filesystem %s: %s", daemon.driver, container.ID, err)
@@ -286,10 +302,6 @@ func (daemon *Daemon) Destroy(container *Container) error {
return fmt.Errorf("Driver %s failed to remove init filesystem %s: %s", daemon.driver, initID, err)
}
if _, err := daemon.containerGraph.Purge(container.ID); err != nil {
utils.Debugf("Unable to remove container from link graph: %s", err)
}
if err := os.RemoveAll(container.root); err != nil {
return fmt.Errorf("Unable to remove filesystem for %v: %v", container.ID, err)
}
@@ -299,20 +311,25 @@ func (daemon *Daemon) Destroy(container *Container) error {
}
func (daemon *Daemon) restore() error {
if os.Getenv("DEBUG") == "" && os.Getenv("TEST") == "" {
var (
debug = (os.Getenv("DEBUG") != "" || os.Getenv("TEST") != "")
containers = make(map[string]*Container)
currentDriver = daemon.driver.String()
containersToStart = []*Container{}
)
if !debug {
fmt.Printf("Loading containers: ")
}
dir, err := ioutil.ReadDir(daemon.repository)
if err != nil {
return err
}
containers := make(map[string]*Container)
currentDriver := daemon.driver.String()
for _, v := range dir {
id := v.Name()
container, err := daemon.load(id)
if os.Getenv("DEBUG") == "" && os.Getenv("TEST") == "" {
if !debug {
fmt.Print(".")
}
if err != nil {
@@ -329,20 +346,16 @@ func (daemon *Daemon) restore() error {
}
}
register := func(container *Container) {
if err := daemon.Register(container); err != nil {
utils.Debugf("Failed to register container %s: %s", container.ID, err)
}
}
if entities := daemon.containerGraph.List("/", -1); entities != nil {
for _, p := range entities.Paths() {
if os.Getenv("DEBUG") == "" && os.Getenv("TEST") == "" {
if !debug {
fmt.Print(".")
}
e := entities[p]
if container, ok := containers[e.ID()]; ok {
register(container)
if err := daemon.register(container, false, &containersToStart); err != nil {
utils.Debugf("Failed to register container %s: %s", container.ID, err)
}
delete(containers, e.ID())
}
}
@@ -351,18 +364,25 @@ func (daemon *Daemon) restore() error {
// Any containers that are left over do not exist in the graph
for _, container := range containers {
// Try to set the default name for a container if it exists prior to links
container.Name, err = generateRandomName(daemon)
container.Name, err = daemon.generateNewName(container.ID)
if err != nil {
container.Name = utils.TruncateID(container.ID)
}
if _, err := daemon.containerGraph.Set(container.Name, container.ID); err != nil {
utils.Debugf("Setting default id - %s", err)
}
register(container)
if err := daemon.register(container, false, &containersToStart); err != nil {
utils.Debugf("Failed to register container %s: %s", container.ID, err)
}
}
if os.Getenv("DEBUG") == "" && os.Getenv("TEST") == "" {
daemon.idIndex.UpdateSuffixarray()
for _, container := range containersToStart {
utils.Debugf("Starting container %d", container.ID)
if err := container.Start(); err != nil {
utils.Debugf("Failed to start container %s: %s", container.ID, err)
}
}
if !debug {
fmt.Printf(": done.\n")
}
@@ -450,42 +470,75 @@ func (daemon *Daemon) generateIdAndName(name string) (string, string, error) {
)
if name == "" {
name, err = generateRandomName(daemon)
if err != nil {
name = utils.TruncateID(id)
}
} else {
if !validContainerNamePattern.MatchString(name) {
return "", "", fmt.Errorf("Invalid container name (%s), only %s are allowed", name, validContainerNameChars)
if name, err = daemon.generateNewName(id); err != nil {
return "", "", err
}
return id, name, nil
}
if name, err = daemon.reserveName(id, name); err != nil {
return "", "", err
}
return id, name, nil
}
func (daemon *Daemon) reserveName(id, name string) (string, error) {
if !validContainerNamePattern.MatchString(name) {
return "", fmt.Errorf("Invalid container name (%s), only %s are allowed", name, validContainerNameChars)
}
if name[0] != '/' {
name = "/" + name
}
// Set the enitity in the graph using the default name specified
if _, err := daemon.containerGraph.Set(name, id); err != nil {
if !graphdb.IsNonUniqueNameError(err) {
return "", "", err
return "", err
}
conflictingContainer, err := daemon.GetByName(name)
if err != nil {
if strings.Contains(err.Error(), "Could not find entity") {
return "", "", err
return "", err
}
// Remove name and continue starting the container
if err := daemon.containerGraph.Delete(name); err != nil {
return "", "", err
return "", err
}
} else {
nameAsKnownByUser := strings.TrimPrefix(name, "/")
return "", "", fmt.Errorf(
return "", fmt.Errorf(
"Conflict, The name %s is already assigned to %s. You have to delete (or rename) that container to be able to assign %s to a container again.", nameAsKnownByUser,
utils.TruncateID(conflictingContainer.ID), nameAsKnownByUser)
}
}
return id, name, nil
return name, nil
}
func (daemon *Daemon) generateNewName(id string) (string, error) {
var name string
for i := 0; i < 6; i++ {
name = namesgenerator.GetRandomName(i)
if name[0] != '/' {
name = "/" + name
}
if _, err := daemon.containerGraph.Set(name, id); err != nil {
if !graphdb.IsNonUniqueNameError(err) {
return "", err
}
continue
}
return name, nil
}
name = "/" + utils.TruncateID(id)
if _, err := daemon.containerGraph.Set(name, id); err != nil {
return "", err
}
return name, nil
}
func (daemon *Daemon) generateHostname(id string, config *runconfig.Config) {
@@ -592,15 +645,18 @@ func (daemon *Daemon) Commit(container *Container, repository, tag, comment, aut
containerID, containerImage string
containerConfig *runconfig.Config
)
if container != nil {
containerID = container.ID
containerImage = container.Image
containerConfig = container.Config
}
img, err := daemon.graph.Create(rwTar, containerID, containerImage, comment, author, containerConfig, config)
if err != nil {
return nil, err
}
// Register the image if needed
if repository != "" {
if err := daemon.repositories.Set(repository, tag, img.ID, true); err != nil {
@@ -629,11 +685,11 @@ func (daemon *Daemon) GetByName(name string) (*Container, error) {
if entity == nil {
return nil, fmt.Errorf("Could not find entity for %s", name)
}
e := daemon.getContainerElement(entity.ID())
e := daemon.containers.Get(entity.ID())
if e == nil {
return nil, fmt.Errorf("Could not find container for entity id %s", entity.ID())
}
return e.Value.(*Container), nil
return e, nil
}
func (daemon *Daemon) Children(name string) (map[string]*Container, error) {
@@ -667,6 +723,35 @@ func (daemon *Daemon) RegisterLink(parent, child *Container, alias string) error
return nil
}
func (daemon *Daemon) RegisterLinks(container *Container, hostConfig *runconfig.HostConfig) error {
if hostConfig != nil && hostConfig.Links != nil {
for _, l := range hostConfig.Links {
parts, err := utils.PartParser("name:alias", l)
if err != nil {
return err
}
child, err := daemon.GetByName(parts["name"])
if err != nil {
return err
}
if child == nil {
return fmt.Errorf("Could not get container for %s", parts["name"])
}
if err := daemon.RegisterLink(container, child, parts["alias"]); err != nil {
return err
}
}
// After we load all the links into the daemon
// set them to nil on the hostconfig
hostConfig.Links = nil
if err := container.WriteHostConfig(); err != nil {
return err
}
}
return nil
}
// FIXME: harmonize with NewGraph()
func NewDaemon(config *daemonconfig.Config, eng *engine.Engine) (*Daemon, error) {
daemon, err := NewDaemonFromDirectory(config, eng)
@@ -680,20 +765,22 @@ func NewDaemonFromDirectory(config *daemonconfig.Config, eng *engine.Engine) (*D
if !config.EnableSelinuxSupport {
selinux.SetDisabled()
}
// Create the root directory if it doesn't exists
if err := os.MkdirAll(config.Root, 0700); err != nil && !os.IsExist(err) {
return nil, err
}
// Set the default driver
graphdriver.DefaultDriver = config.GraphDriver
// Load storage driver
driver, err := graphdriver.New(config.Root)
driver, err := graphdriver.New(config.Root, config.GraphOptions)
if err != nil {
return nil, err
}
utils.Debugf("Using graph driver %s", driver)
if err := remountPrivate(config.Root); err != nil {
return nil, err
}
daemonRepo := path.Join(config.Root, "containers")
if err := os.MkdirAll(daemonRepo, 0700); err != nil && !os.IsExist(err) {
@@ -713,7 +800,7 @@ func NewDaemonFromDirectory(config *daemonconfig.Config, eng *engine.Engine) (*D
// We don't want to use a complex driver like aufs or devmapper
// for volumes, just a plain filesystem
volumesDriver, err := graphdriver.GetDriver("vfs", config.Root)
volumesDriver, err := graphdriver.GetDriver("vfs", config.Root, config.GraphOptions)
if err != nil {
return nil, err
}
@@ -777,7 +864,7 @@ func NewDaemonFromDirectory(config *daemonconfig.Config, eng *engine.Engine) (*D
daemon := &Daemon{
repository: daemonRepo,
containers: list.New(),
containers: &contStore{s: make(map[string]*Container)},
graph: g,
repositories: repositories,
idIndex: utils.NewTruncIndex([]string{}),
@@ -914,6 +1001,22 @@ func (daemon *Daemon) Run(c *Container, pipes *execdriver.Pipes, startCallback e
return daemon.execDriver.Run(c.command, pipes, startCallback)
}
func (daemon *Daemon) Pause(c *Container) error {
if err := daemon.execDriver.Pause(c.command); err != nil {
return err
}
c.State.SetPaused()
return nil
}
func (daemon *Daemon) Unpause(c *Container) error {
if err := daemon.execDriver.Unpause(c.command); err != nil {
return err
}
c.State.SetUnpaused()
return nil
}
func (daemon *Daemon) Kill(c *Container, sig int) error {
return daemon.execDriver.Kill(c.command, sig)
}
+24 -17
View File
@@ -5,6 +5,8 @@ import (
"io"
"os"
"os/exec"
"github.com/dotcloud/docker/pkg/libcontainer/devices"
)
// Context is a generic key value pair that allows
@@ -81,6 +83,8 @@ type TtyTerminal interface {
type Driver interface {
Run(c *Command, pipes *Pipes, startCallback StartCallback) (int, error) // Run executes the process and blocks until the process exits and returns the exit code
Kill(c *Command, sig int) error
Pause(c *Command) error
Unpause(c *Command) error
Name() string // Driver name
Info(id string) Info // "temporary" hack (until we move state from core to plugins)
GetPidsForContainer(id string) ([]int, error) // Returns a list of pids for the given container.
@@ -103,9 +107,10 @@ type NetworkInterface struct {
}
type Resources struct {
Memory int64 `json:"memory"`
MemorySwap int64 `json:"memory_swap"`
CpuShares int64 `json:"cpu_shares"`
Memory int64 `json:"memory"`
MemorySwap int64 `json:"memory_swap"`
CpuShares int64 `json:"cpu_shares"`
Cpuset string `json:"cpuset"`
}
type Mount struct {
@@ -119,20 +124,22 @@ type Mount struct {
type Command struct {
exec.Cmd `json:"-"`
ID string `json:"id"`
Privileged bool `json:"privileged"`
User string `json:"user"`
Rootfs string `json:"rootfs"` // root fs of the container
InitPath string `json:"initpath"` // dockerinit
Entrypoint string `json:"entrypoint"`
Arguments []string `json:"arguments"`
WorkingDir string `json:"working_dir"`
ConfigPath string `json:"config_path"` // this should be able to be removed when the lxc template is moved into the driver
Tty bool `json:"tty"`
Network *Network `json:"network"`
Config map[string][]string `json:"config"` // generic values that specific drivers can consume
Resources *Resources `json:"resources"`
Mounts []Mount `json:"mounts"`
ID string `json:"id"`
Privileged bool `json:"privileged"`
User string `json:"user"`
Rootfs string `json:"rootfs"` // root fs of the container
InitPath string `json:"initpath"` // dockerinit
Entrypoint string `json:"entrypoint"`
Arguments []string `json:"arguments"`
WorkingDir string `json:"working_dir"`
ConfigPath string `json:"config_path"` // this should be able to be removed when the lxc template is moved into the driver
Tty bool `json:"tty"`
Network *Network `json:"network"`
Config map[string][]string `json:"config"` // generic values that specific drivers can consume
Resources *Resources `json:"resources"`
Mounts []Mount `json:"mounts"`
AllowedDevices []*devices.Device `json:"allowed_devices"`
AutoCreatedDevices []*devices.Device `json:"autocreated_devices"`
Terminal Terminal `json:"-"` // standard or tty terminal
Console string `json:"-"` // dev/console path
+1 -1
View File
@@ -12,7 +12,7 @@ import (
func NewDriver(name, root, initPath string, sysInfo *sysinfo.SysInfo) (execdriver.Driver, error) {
switch name {
case "lxc":
// we want to five the lxc driver the full docker root because it needs
// we want to give the lxc driver the full docker root because it needs
// to access and write config and template files in /var/lib/docker/containers/*
// to be backwards compatible
return lxc.NewDriver(root, sysInfo.AppArmor)
+40 -10
View File
@@ -9,14 +9,16 @@ import (
"os/exec"
"path"
"path/filepath"
"runtime"
"strconv"
"strings"
"syscall"
"time"
"github.com/dotcloud/docker/daemon/execdriver"
"github.com/dotcloud/docker/pkg/cgroups"
"github.com/dotcloud/docker/pkg/label"
"github.com/dotcloud/docker/pkg/libcontainer/cgroups"
"github.com/dotcloud/docker/pkg/libcontainer/mount/nodes"
"github.com/dotcloud/docker/pkg/system"
"github.com/dotcloud/docker/utils"
)
@@ -25,6 +27,7 @@ const DriverName = "lxc"
func init() {
execdriver.RegisterInitFunc(DriverName, func(args *execdriver.InitArgs) error {
runtime.LockOSThread()
if err := setupEnv(args); err != nil {
return err
}
@@ -159,6 +162,10 @@ func (d *driver) Run(c *execdriver.Command, pipes *execdriver.Pipes, startCallba
c.Path = aname
c.Args = append([]string{name}, arg...)
if err := nodes.CreateDeviceNodes(c.Rootfs, c.AutoCreatedDevices); err != nil {
return -1, err
}
if err := c.Start(); err != nil {
return -1, err
}
@@ -167,6 +174,7 @@ func (d *driver) Run(c *execdriver.Command, pipes *execdriver.Pipes, startCallba
waitErr error
waitLock = make(chan struct{})
)
go func() {
if err := c.Wait(); err != nil {
if _, ok := err.(*exec.ExitError); !ok { // Do not propagate the error if it's simply a status code != 0
@@ -181,9 +189,11 @@ func (d *driver) Run(c *execdriver.Command, pipes *execdriver.Pipes, startCallba
if err != nil {
if c.Process != nil {
c.Process.Kill()
c.Wait()
}
return -1, err
}
c.ContainerPid = pid
if startCallback != nil {
@@ -208,6 +218,30 @@ func (d *driver) Kill(c *execdriver.Command, sig int) error {
return KillLxc(c.ID, sig)
}
func (d *driver) Pause(c *execdriver.Command) error {
_, err := exec.LookPath("lxc-freeze")
if err == nil {
output, errExec := exec.Command("lxc-freeze", "-n", c.ID).CombinedOutput()
if errExec != nil {
return fmt.Errorf("Err: %s Output: %s", errExec, output)
}
}
return err
}
func (d *driver) Unpause(c *execdriver.Command) error {
_, err := exec.LookPath("lxc-unfreeze")
if err == nil {
output, errExec := exec.Command("lxc-unfreeze", "-n", c.ID).CombinedOutput()
if errExec != nil {
return fmt.Errorf("Err: %s Output: %s", errExec, output)
}
}
return err
}
func (d *driver) Terminate(c *execdriver.Command) error {
return KillLxc(c.ID, 9)
}
@@ -268,18 +302,14 @@ func (d *driver) waitForStart(c *execdriver.Command, waitLock chan struct{}) (in
}
output, err = d.getInfo(c.ID)
if err != nil {
output, err = d.getInfo(c.ID)
if err == nil {
info, err := parseLxcInfo(string(output))
if err != nil {
return -1, err
}
}
info, err := parseLxcInfo(string(output))
if err != nil {
return -1, err
}
if info.Running {
return info.Pid, nil
if info.Running {
return info.Pid, nil
}
}
time.Sleep(50 * time.Millisecond)
}
+1 -1
View File
@@ -88,7 +88,7 @@ func setupNetworking(args *execdriver.InitArgs) error {
return fmt.Errorf("Unable to set up networking, %s is not a valid gateway IP", args.Gateway)
}
if err := netlink.AddDefaultGw(gw); err != nil {
if err := netlink.AddDefaultGw(gw.String(), "eth0"); err != nil {
return fmt.Errorf("Unable to set up networking: %v", err)
}
}
+10 -32
View File
@@ -15,7 +15,9 @@ lxc.network.type = veth
lxc.network.link = {{.Network.Interface.Bridge}}
lxc.network.name = eth0
lxc.network.mtu = {{.Network.Mtu}}
{{else if not .Network.HostNetworking}}
{{else if .Network.HostNetworking}}
lxc.network.type = none
{{else}}
# network is disabled (-n=false)
lxc.network.type = empty
lxc.network.flags = up
@@ -45,37 +47,10 @@ lxc.cgroup.devices.allow = a
{{else}}
# no implicit access to devices
lxc.cgroup.devices.deny = a
# but allow mknod for any device
lxc.cgroup.devices.allow = c *:* m
lxc.cgroup.devices.allow = b *:* m
# /dev/null and zero
lxc.cgroup.devices.allow = c 1:3 rwm
lxc.cgroup.devices.allow = c 1:5 rwm
# consoles
lxc.cgroup.devices.allow = c 5:1 rwm
lxc.cgroup.devices.allow = c 5:0 rwm
lxc.cgroup.devices.allow = c 4:0 rwm
lxc.cgroup.devices.allow = c 4:1 rwm
# /dev/urandom,/dev/random
lxc.cgroup.devices.allow = c 1:9 rwm
lxc.cgroup.devices.allow = c 1:8 rwm
# /dev/pts/ - pts namespaces are "coming soon"
lxc.cgroup.devices.allow = c 136:* rwm
lxc.cgroup.devices.allow = c 5:2 rwm
# tuntap
lxc.cgroup.devices.allow = c 10:200 rwm
# fuse
#lxc.cgroup.devices.allow = c 10:229 rwm
# rtc
#lxc.cgroup.devices.allow = c 254:0 rwm
#Allow the devices passed to us in the AllowedDevices list.
{{range $allowedDevice := .AllowedDevices}}
lxc.cgroup.devices.allow = {{$allowedDevice.GetCgroupAllowString}}
{{end}}
{{end}}
# standard mount point
@@ -126,6 +101,9 @@ lxc.cgroup.memory.memsw.limit_in_bytes = {{$memSwap}}
{{if .Resources.CpuShares}}
lxc.cgroup.cpu.shares = {{.Resources.CpuShares}}
{{end}}
{{if .Resources.Cpuset}}
lxc.cgroup.cpuset.cpus = {{.Resources.Cpuset}}
{{end}}
{{end}}
{{if .Config.lxc}}
@@ -3,7 +3,6 @@ package lxc
import (
"bufio"
"fmt"
"github.com/dotcloud/docker/daemon/execdriver"
"io/ioutil"
"math/rand"
"os"
@@ -11,6 +10,9 @@ import (
"strings"
"testing"
"time"
"github.com/dotcloud/docker/daemon/execdriver"
"github.com/dotcloud/docker/pkg/libcontainer/devices"
)
func TestLXCConfig(t *testing.T) {
@@ -47,6 +49,7 @@ func TestLXCConfig(t *testing.T) {
Mtu: 1500,
Interface: nil,
},
AllowedDevices: make([]*devices.Device, 0),
}
p, err := driver.generateLXCConfig(command)
if err != nil {
@@ -8,7 +8,7 @@ import (
"strings"
"github.com/dotcloud/docker/pkg/libcontainer"
"github.com/dotcloud/docker/utils"
"github.com/dotcloud/docker/pkg/units"
)
type Action func(*libcontainer.Container, interface{}, string) error
@@ -75,7 +75,7 @@ func memory(container *libcontainer.Container, context interface{}, value string
return fmt.Errorf("cannot set cgroups when they are disabled")
}
v, err := utils.RAMInBytes(value)
v, err := units.RAMInBytes(value)
if err != nil {
return err
}
@@ -88,7 +88,7 @@ func memoryReservation(container *libcontainer.Container, context interface{}, v
return fmt.Errorf("cannot set cgroups when they are disabled")
}
v, err := utils.RAMInBytes(value)
v, err := units.RAMInBytes(value)
if err != nil {
return err
}
@@ -109,12 +109,19 @@ func memorySwap(container *libcontainer.Container, context interface{}, value st
}
func addCap(container *libcontainer.Container, context interface{}, value string) error {
container.CapabilitiesMask[value] = true
container.Capabilities = append(container.Capabilities, value)
return nil
}
func dropCap(container *libcontainer.Container, context interface{}, value string) error {
container.CapabilitiesMask[value] = false
// If the capability is specified multiple times, remove all instances.
for i, capability := range container.Capabilities {
if capability == value {
container.Capabilities = append(container.Capabilities[:i], container.Capabilities[i+1:]...)
}
}
// The capability wasn't found so we will drop it anyways.
return nil
}
@@ -4,8 +4,19 @@ import (
"testing"
"github.com/dotcloud/docker/daemon/execdriver/native/template"
"github.com/dotcloud/docker/pkg/libcontainer"
)
// Checks whether the expected capability is specified in the capabilities.
func hasCapability(expected string, capabilities []string) bool {
for _, capability := range capabilities {
if capability == expected {
return true
}
}
return false
}
func TestSetReadonlyRootFs(t *testing.T) {
var (
container = template.New()
@@ -39,10 +50,10 @@ func TestConfigurationsDoNotConflict(t *testing.T) {
t.Fatal(err)
}
if !container1.CapabilitiesMask["NET_ADMIN"] {
if !hasCapability("NET_ADMIN", container1.Capabilities) {
t.Fatal("container one should have NET_ADMIN enabled")
}
if container2.CapabilitiesMask["NET_ADMIN"] {
if hasCapability("NET_ADMIN", container2.Capabilities) {
t.Fatal("container two should not have NET_ADMIN enabled")
}
}
@@ -138,10 +149,10 @@ func TestAddCap(t *testing.T) {
t.Fatal(err)
}
if !container.CapabilitiesMask["MKNOD"] {
if !hasCapability("MKNOD", container.Capabilities) {
t.Fatal("container should have MKNOD enabled")
}
if !container.CapabilitiesMask["SYS_ADMIN"] {
if !hasCapability("SYS_ADMIN", container.Capabilities) {
t.Fatal("container should have SYS_ADMIN enabled")
}
}
@@ -154,14 +165,12 @@ func TestDropCap(t *testing.T) {
}
)
// enabled all caps like in privileged mode
for key := range container.CapabilitiesMask {
container.CapabilitiesMask[key] = true
}
container.Capabilities = libcontainer.GetAllCapabilities()
if err := ParseConfiguration(container, nil, opts); err != nil {
t.Fatal(err)
}
if container.CapabilitiesMask["MKNOD"] {
if hasCapability("MKNOD", container.Capabilities) {
t.Fatal("container should not have MKNOD enabled")
}
}
+26 -9
View File
@@ -3,6 +3,7 @@ package native
import (
"fmt"
"os"
"os/exec"
"path/filepath"
"github.com/dotcloud/docker/daemon/execdriver"
@@ -10,6 +11,7 @@ import (
"github.com/dotcloud/docker/daemon/execdriver/native/template"
"github.com/dotcloud/docker/pkg/apparmor"
"github.com/dotcloud/docker/pkg/libcontainer"
"github.com/dotcloud/docker/pkg/libcontainer/devices"
)
// createContainer populates and configures the container type with the
@@ -23,6 +25,8 @@ func (d *driver) createContainer(c *execdriver.Command) (*libcontainer.Container
container.WorkingDir = c.WorkingDir
container.Env = c.Env
container.Cgroups.Name = c.ID
container.Cgroups.AllowedDevices = c.AllowedDevices
container.DeviceNodes = c.AutoCreatedDevices
// check to see if we are running in ramdisk to disable pivot root
container.NoPivotRoot = os.Getenv("DOCKER_RAMDISK") != ""
container.Context["restrictions"] = "true"
@@ -34,8 +38,6 @@ func (d *driver) createContainer(c *execdriver.Command) (*libcontainer.Container
if err := d.setPrivileged(container); err != nil {
return nil, err
}
} else {
container.Mounts = append(container.Mounts, libcontainer.Mount{Type: "devtmpfs"})
}
if err := d.setupCgroups(container, c); err != nil {
return nil, err
@@ -46,7 +48,13 @@ func (d *driver) createContainer(c *execdriver.Command) (*libcontainer.Container
if err := d.setupLabels(container, c); err != nil {
return nil, err
}
if err := configuration.ParseConfiguration(container, d.activeContainers, c.Config["native"]); err != nil {
cmds := make(map[string]*exec.Cmd)
d.Lock()
for k, v := range d.activeContainers {
cmds[k] = v.cmd
}
d.Unlock()
if err := configuration.ParseConfiguration(container, cmds, c.Config["native"]); err != nil {
return nil, err
}
return container, nil
@@ -82,10 +90,14 @@ func (d *driver) createNetwork(container *libcontainer.Container, c *execdriver.
}
if c.Network.ContainerID != "" {
cmd := d.activeContainers[c.Network.ContainerID]
if cmd == nil || cmd.Process == nil {
d.Lock()
active := d.activeContainers[c.Network.ContainerID]
d.Unlock()
if active == nil || active.cmd.Process == nil {
return fmt.Errorf("%s is not a valid running container to join", c.Network.ContainerID)
}
cmd := active.cmd
nspath := filepath.Join("/proc", fmt.Sprint(cmd.Process.Pid), "ns", "net")
container.Networks = append(container.Networks, &libcontainer.Network{
Type: "netns",
@@ -97,11 +109,15 @@ func (d *driver) createNetwork(container *libcontainer.Container, c *execdriver.
return nil
}
func (d *driver) setPrivileged(container *libcontainer.Container) error {
for key := range container.CapabilitiesMask {
container.CapabilitiesMask[key] = true
func (d *driver) setPrivileged(container *libcontainer.Container) (err error) {
container.Capabilities = libcontainer.GetAllCapabilities()
container.Cgroups.AllowAllDevices = true
hostDeviceNodes, err := devices.GetHostDeviceNodes()
if err != nil {
return err
}
container.Cgroups.DeviceAccess = true
container.DeviceNodes = hostDeviceNodes
delete(container.Context, "restrictions")
@@ -117,6 +133,7 @@ func (d *driver) setupCgroups(container *libcontainer.Container, c *execdriver.C
container.Cgroups.Memory = c.Resources.Memory
container.Cgroups.MemoryReservation = c.Resources.Memory
container.Cgroups.MemorySwap = c.Resources.MemorySwap
container.Cgroups.CpusetCpus = c.Resources.Cpuset
}
return nil
}
+68 -47
View File
@@ -7,22 +7,22 @@ import (
"os"
"os/exec"
"path/filepath"
"strconv"
"strings"
"sync"
"syscall"
"github.com/dotcloud/docker/daemon/execdriver"
"github.com/dotcloud/docker/pkg/apparmor"
"github.com/dotcloud/docker/pkg/cgroups"
"github.com/dotcloud/docker/pkg/libcontainer"
"github.com/dotcloud/docker/pkg/libcontainer/nsinit"
"github.com/dotcloud/docker/pkg/libcontainer/cgroups/fs"
"github.com/dotcloud/docker/pkg/libcontainer/cgroups/systemd"
"github.com/dotcloud/docker/pkg/libcontainer/namespaces"
"github.com/dotcloud/docker/pkg/system"
)
const (
DriverName = "native"
Version = "0.2"
BackupApparmorProfilePath = "apparmor/docker.back" // relative to docker root
DriverName = "native"
Version = "0.2"
)
func init() {
@@ -42,35 +42,43 @@ func init() {
if err != nil {
return err
}
syncPipe, err := nsinit.NewSyncPipeFromFd(0, uintptr(args.Pipe))
syncPipe, err := namespaces.NewSyncPipeFromFd(0, uintptr(args.Pipe))
if err != nil {
return err
}
if err := nsinit.Init(container, rootfs, args.Console, syncPipe, args.Args); err != nil {
if err := namespaces.Init(container, rootfs, args.Console, syncPipe, args.Args); err != nil {
return err
}
return nil
})
}
type activeContainer struct {
container *libcontainer.Container
cmd *exec.Cmd
}
type driver struct {
root string
initPath string
activeContainers map[string]*exec.Cmd
activeContainers map[string]*activeContainer
sync.Mutex
}
func NewDriver(root, initPath string) (*driver, error) {
if err := os.MkdirAll(root, 0700); err != nil {
return nil, err
}
// native driver root is at docker_root/execdriver/native. Put apparmor at docker_root
if err := apparmor.InstallDefaultProfile(filepath.Join(root, "../..", BackupApparmorProfilePath)); err != nil {
if err := apparmor.InstallDefaultProfile(); err != nil {
return nil, err
}
return &driver{
root: root,
initPath: initPath,
activeContainers: make(map[string]*exec.Cmd),
activeContainers: make(map[string]*activeContainer),
}, nil
}
@@ -80,7 +88,12 @@ func (d *driver) Run(c *execdriver.Command, pipes *execdriver.Pipes, startCallba
if err != nil {
return -1, err
}
d.activeContainers[c.ID] = &c.Cmd
d.Lock()
d.activeContainers[c.ID] = &activeContainer{
container: container,
cmd: &c.Cmd,
}
d.Unlock()
var (
dataPath = filepath.Join(d.root, c.ID)
@@ -97,8 +110,8 @@ func (d *driver) Run(c *execdriver.Command, pipes *execdriver.Pipes, startCallba
term := getTerminal(c, pipes)
return nsinit.Exec(container, term, c.Rootfs, dataPath, args, func(container *libcontainer.Container, console, rootfs, dataPath, init string, child *os.File, args []string) *exec.Cmd {
// we need to join the rootfs because nsinit will setup the rootfs and chroot
return namespaces.Exec(container, term, c.Rootfs, dataPath, args, func(container *libcontainer.Container, console, rootfs, dataPath, init string, child *os.File, args []string) *exec.Cmd {
// we need to join the rootfs because namespaces will setup the rootfs and chroot
initPath := filepath.Join(c.Rootfs, c.InitPath)
c.Path = d.initPath
@@ -113,7 +126,7 @@ func (d *driver) Run(c *execdriver.Command, pipes *execdriver.Pipes, startCallba
// set this to nil so that when we set the clone flags anything else is reset
c.SysProcAttr = nil
system.SetCloneFlags(&c.Cmd, uintptr(nsinit.GetNamespaceFlags(container.Namespaces)))
system.SetCloneFlags(&c.Cmd, uintptr(namespaces.GetNamespaceFlags(container.Namespaces)))
c.ExtraFiles = []*os.File{child}
c.Env = container.Env
@@ -132,6 +145,30 @@ func (d *driver) Kill(p *execdriver.Command, sig int) error {
return syscall.Kill(p.Process.Pid, syscall.Signal(sig))
}
func (d *driver) Pause(c *execdriver.Command) error {
active := d.activeContainers[c.ID]
if active == nil {
return fmt.Errorf("active container for %s does not exist", c.ID)
}
active.container.Cgroups.Freezer = "FROZEN"
if systemd.UseSystemd() {
return systemd.Freeze(active.container.Cgroups, active.container.Cgroups.Freezer)
}
return fs.Freeze(active.container.Cgroups, active.container.Cgroups.Freezer)
}
func (d *driver) Unpause(c *execdriver.Command) error {
active := d.activeContainers[c.ID]
if active == nil {
return fmt.Errorf("active container for %s does not exist", c.ID)
}
active.container.Cgroups.Freezer = "THAWED"
if systemd.UseSystemd() {
return systemd.Freeze(active.container.Cgroups, active.container.Cgroups.Freezer)
}
return fs.Freeze(active.container.Cgroups, active.container.Cgroups.Freezer)
}
func (d *driver) Terminate(p *execdriver.Command) error {
// lets check the start time for the process
started, err := d.readStartTime(p)
@@ -150,6 +187,7 @@ func (d *driver) Terminate(p *execdriver.Command) error {
}
if started == currentStartTime {
err = syscall.Kill(p.Process.Pid, 9)
syscall.Wait4(p.Process.Pid, nil, 0, nil)
}
d.removeContainerRoot(p.ID)
return err
@@ -175,41 +213,20 @@ func (d *driver) Name() string {
return fmt.Sprintf("%s-%s", DriverName, Version)
}
// TODO: this can be improved with our driver
// there has to be a better way to do this
func (d *driver) GetPidsForContainer(id string) ([]int, error) {
pids := []int{}
d.Lock()
active := d.activeContainers[id]
d.Unlock()
subsystem := "devices"
cgroupRoot, err := cgroups.FindCgroupMountpoint(subsystem)
if err != nil {
return pids, err
}
cgroupDir, err := cgroups.GetThisCgroupDir(subsystem)
if err != nil {
return pids, err
if active == nil {
return nil, fmt.Errorf("active container for %s does not exist", id)
}
c := active.container.Cgroups
filename := filepath.Join(cgroupRoot, cgroupDir, id, "tasks")
if _, err := os.Stat(filename); os.IsNotExist(err) {
filename = filepath.Join(cgroupRoot, cgroupDir, "docker", id, "tasks")
if systemd.UseSystemd() {
return systemd.GetPids(c)
}
output, err := ioutil.ReadFile(filename)
if err != nil {
return pids, err
}
for _, p := range strings.Split(string(output), "\n") {
if len(p) == 0 {
continue
}
pid, err := strconv.Atoi(p)
if err != nil {
return pids, fmt.Errorf("Invalid pid '%s': %s", p, err)
}
pids = append(pids, pid)
}
return pids, nil
return fs.GetPids(c)
}
func (d *driver) writeContainerFile(container *libcontainer.Container, id string) error {
@@ -225,6 +242,10 @@ func (d *driver) createContainerRoot(id string) error {
}
func (d *driver) removeContainerRoot(id string) error {
d.Lock()
delete(d.activeContainers, id)
d.Unlock()
return os.RemoveAll(filepath.Join(d.root, id))
}
@@ -238,8 +259,8 @@ func getEnv(key string, env []string) string {
return ""
}
func getTerminal(c *execdriver.Command, pipes *execdriver.Pipes) nsinit.Terminal {
var term nsinit.Terminal
func getTerminal(c *execdriver.Command, pipes *execdriver.Pipes) namespaces.Terminal {
var term namespaces.Terminal
if c.Tty {
term = &dockerTtyTerm{
pipes: pipes,
@@ -2,30 +2,25 @@ package template
import (
"github.com/dotcloud/docker/pkg/apparmor"
"github.com/dotcloud/docker/pkg/cgroups"
"github.com/dotcloud/docker/pkg/libcontainer"
"github.com/dotcloud/docker/pkg/libcontainer/cgroups"
)
// New returns the docker default configuration for libcontainer
func New() *libcontainer.Container {
container := &libcontainer.Container{
CapabilitiesMask: map[string]bool{
"SETPCAP": false,
"SYS_MODULE": false,
"SYS_RAWIO": false,
"SYS_PACCT": false,
"SYS_ADMIN": false,
"SYS_NICE": false,
"SYS_RESOURCE": false,
"SYS_TIME": false,
"SYS_TTY_CONFIG": false,
"AUDIT_WRITE": false,
"AUDIT_CONTROL": false,
"MAC_OVERRIDE": false,
"MAC_ADMIN": false,
"NET_ADMIN": false,
"MKNOD": true,
"SYSLOG": false,
Capabilities: []string{
"CHOWN",
"DAC_OVERRIDE",
"FOWNER",
"MKNOD",
"NET_RAW",
"SETGID",
"SETUID",
"SETFCAP",
"SETPCAP",
"NET_BIND_SERVICE",
"SYS_CHROOT",
},
Namespaces: map[string]bool{
"NEWNS": true,
@@ -35,8 +30,8 @@ func New() *libcontainer.Container {
"NEWNET": true,
},
Cgroups: &cgroups.Cgroup{
Parent: "docker",
DeviceAccess: false,
Parent: "docker",
AllowAllDevices: false,
},
Context: libcontainer.Context{},
}
+35 -9
View File
@@ -7,7 +7,7 @@ aufs driver directory structure
   1
   2
   3
diffs // Content of the layer
diff // Content of the layer
   1 // Contains layers that need to be mounted for the id
   2
   3
@@ -23,20 +23,26 @@ package aufs
import (
"bufio"
"fmt"
"github.com/dotcloud/docker/archive"
"github.com/dotcloud/docker/daemon/graphdriver"
"github.com/dotcloud/docker/pkg/label"
mountpk "github.com/dotcloud/docker/pkg/mount"
"github.com/dotcloud/docker/utils"
"os"
"os/exec"
"path"
"strings"
"sync"
"syscall"
"github.com/dotcloud/docker/archive"
"github.com/dotcloud/docker/daemon/graphdriver"
"github.com/dotcloud/docker/pkg/label"
mountpk "github.com/dotcloud/docker/pkg/mount"
"github.com/dotcloud/docker/utils"
)
var (
ErrAufsNotSupported = fmt.Errorf("AUFS was not found in /proc/filesystems")
incompatibleFsMagic = []graphdriver.FsMagic{
graphdriver.FsMagicBtrfs,
graphdriver.FsMagicAufs,
}
)
func init() {
@@ -51,11 +57,25 @@ type Driver struct {
// New returns a new AUFS driver.
// An error is returned if AUFS is not supported.
func Init(root string) (graphdriver.Driver, error) {
func Init(root string, options []string) (graphdriver.Driver, error) {
// Try to load the aufs kernel module
if err := supportsAufs(); err != nil {
return nil, err
return nil, graphdriver.ErrNotSupported
}
rootdir := path.Dir(root)
var buf syscall.Statfs_t
if err := syscall.Statfs(rootdir, &buf); err != nil {
return nil, fmt.Errorf("Couldn't stat the root directory: %s", err)
}
for _, magic := range incompatibleFsMagic {
if graphdriver.FsMagic(buf.Type) == magic {
return nil, graphdriver.ErrIncompatibleFS
}
}
paths := []string{
"mnt",
"diff",
@@ -77,6 +97,10 @@ func Init(root string) (graphdriver.Driver, error) {
return nil, err
}
if err := graphdriver.MakePrivate(root); err != nil {
return nil, err
}
for _, p := range paths {
if err := os.MkdirAll(path.Join(root, p), 0755); err != nil {
return nil, err
@@ -351,12 +375,14 @@ func (a *Driver) Cleanup() error {
if err != nil {
return err
}
for _, id := range ids {
if err := a.unmount(id); err != nil {
utils.Errorf("Unmounting %s: %s", utils.TruncateID(id), err)
}
}
return nil
return mountpk.Unmount(a.root)
}
func (a *Driver) aufsMount(ro []string, rw, target, mountLabel string) (err error) {
+2 -2
View File
@@ -17,9 +17,9 @@ var (
)
func testInit(dir string, t *testing.T) graphdriver.Driver {
d, err := Init(dir)
d, err := Init(dir, nil)
if err != nil {
if err == ErrAufsNotSupported {
if err == graphdriver.ErrNotSupported {
t.Skip(err)
} else {
t.Fatal(err)
+15 -5
View File
@@ -11,18 +11,20 @@ import "C"
import (
"fmt"
"github.com/dotcloud/docker/daemon/graphdriver"
"os"
"path"
"syscall"
"unsafe"
"github.com/dotcloud/docker/daemon/graphdriver"
"github.com/dotcloud/docker/pkg/mount"
)
func init() {
graphdriver.Register("btrfs", Init)
}
func Init(home string) (graphdriver.Driver, error) {
func Init(home string, options []string) (graphdriver.Driver, error) {
rootdir := path.Dir(home)
var buf syscall.Statfs_t
@@ -30,8 +32,16 @@ func Init(home string) (graphdriver.Driver, error) {
return nil, err
}
if buf.Type != 0x9123683E {
return nil, fmt.Errorf("%s is not a btrfs filesystem", rootdir)
if graphdriver.FsMagic(buf.Type) != graphdriver.FsMagicBtrfs {
return nil, graphdriver.ErrPrerequisites
}
if err := os.MkdirAll(home, 0700); err != nil {
return nil, err
}
if err := graphdriver.MakePrivate(home); err != nil {
return nil, err
}
return &Driver{
@@ -52,7 +62,7 @@ func (d *Driver) Status() [][2]string {
}
func (d *Driver) Cleanup() error {
return nil
return mount.Unmount(d.home)
}
func free(p *C.char) {
+28
View File
@@ -0,0 +1,28 @@
package btrfs
import (
"github.com/dotcloud/docker/daemon/graphdriver/graphtest"
"testing"
)
// This avoids creating a new driver for each test if all tests are run
// Make sure to put new tests between TestBtrfsSetup and TestBtrfsTeardown
func TestBtrfsSetup(t *testing.T) {
graphtest.GetDriver(t, "btrfs")
}
func TestBtrfsCreateEmpty(t *testing.T) {
graphtest.DriverTestCreateEmpty(t, "btrfs")
}
func TestBtrfsCreateBase(t *testing.T) {
graphtest.DriverTestCreateBase(t, "btrfs")
}
func TestBtrfsCreateSnap(t *testing.T) {
graphtest.DriverTestCreateSnap(t, "btrfs")
}
func TestBtrfsTeardown(t *testing.T) {
graphtest.PutDriver(t)
}
+143
View File
@@ -0,0 +1,143 @@
## devicemapper - a storage backend based on Device Mapper
### Theory of operation
The device mapper graphdriver uses the device mapper thin provisioning
module (dm-thinp) to implement CoW snapshots. For each devicemapper
graph location (typically `/var/lib/docker/devicemapper`, $graph below)
a thin pool is created based on two block devices, one for data and
one for metadata. By default these block devices are created
automatically by using loopback mounts of automatically creates sparse
files.
The default loopback files used are `$graph/devicemapper/data` and
`$graph/devicemapper/metadata`. Additional metadata required to map
from docker entities to the corresponding devicemapper volumes is
stored in the `$graph/devicemapper/json` file (encoded as Json).
In order to support multiple devicemapper graphs on a system the thin
pool will be named something like: `docker-0:33-19478248-pool`, where
the `0:30` part is the minor/major device nr and `19478248` is the
inode number of the $graph directory.
On the thin pool docker automatically creates a base thin device,
called something like `docker-0:33-19478248-base` of a fixed
size. This is automatically formated on creation and contains just an
empty filesystem. This device is the base of all docker images and
containers. All base images are snapshots of this device and those
images are then in turn used as snapshots for other images and
eventually containers.
### options
The devicemapper backend supports some options that you can specify
when starting the docker daemon using the --storage-opt flags.
This uses the `dm` prefix and would be used somthing like `docker -d --storage-opt dm.foo=bar`.
Here is the list of supported options:
* `dm.basesize`
Specifies the size to use when creating the base device, which
limits the size of images and containers. The default value is
10G. Note, thin devices are inherently "sparse", so a 10G device
which is mostly empty doesn't use 10 GB of space on the
pool. However, the filesystem will use more space for the empty
case the larger the device is.
Example use:
``docker -d --storage-opt dm.basesize=20G``
* `dm.loopdatasize`
Specifies the size to use when creating the loopback file for the
"data" device which is used for the thin pool. The default size is
100G. Note that the file is sparse, so it will not initially take
up this much space.
Example use:
``docker -d --storage-opt dm.loopdatasize=200G``
* `dm.loopmetadatasize`
Specifies the size to use when creating the loopback file for the
"metadadata" device which is used for the thin pool. The default size is
2G. Note that the file is sparse, so it will not initially take
up this much space.
Example use:
``docker -d --storage-opt dm.loopmetadatasize=4G``
* `dm.fs`
Specifies the filesystem type to use for the base device. The supported
options are "ext4" and "xfs". The default is "ext4"
Example use:
``docker -d --storage-opt dm.fs=xfs``
* `dm.mkfsarg`
Specifies extra mkfs arguments to be used when creating the base device.
Example use:
``docker -d --storage-opt "dm.mkfsarg=-O ^has_journal"``
* `dm.mountopt`
Specifies extra mount options used when mounting the thin devices.
Example use:
``docker -d --storage-opt dm.mountopt=nodiscard``
* `dm.datadev`
Specifies a custom blockdevice to use for data for the thin pool.
If using a block device for device mapper storage, ideally both
datadev and metadatadev should be specified to completely avoid
using the loopback device.
Example use:
``docker -d --storage-opt dm.datadev=/dev/sdb1 --storage-opt dm.metadatadev=/dev/sdc1``
* `dm.metadatadev`
Specifies a custom blockdevice to use for metadata for the thin
pool.
For best performance the metadata should be on a different spindle
than the data, or even better on an SSD.
If setting up a new metadata pool it is required to be valid. This
can be achieved by zeroing the first 4k to indicate empty
metadata, like this:
``dd if=/dev/zero of=$metadata_dev bs=4096 count=1```
Example use:
``docker -d --storage-opt dm.datadev=/dev/sdb1 --storage-opt dm.metadatadev=/dev/sdc1``
* `dm.blkdiscard`
Enables or disables the use of blkdiscard when removing
devicemapper devices. This is enabled by default (only) if using
loopback devices and is required to res-parsify the loopback file
on image/container removal.
Disabling this on loopback can lead to *much* faster container
removal times, but will make the space used in /var/lib/docker
directory not be returned to the system for other use when
containers are removed.
Example use:
``docker -d --storage-opt dm.blkdiscard=false``
+13 -10
View File
@@ -4,6 +4,9 @@ package devmapper
import (
"fmt"
"os"
"syscall"
"github.com/dotcloud/docker/utils"
)
@@ -14,7 +17,7 @@ func stringToLoopName(src string) [LoNameSize]uint8 {
}
func getNextFreeLoopbackIndex() (int, error) {
f, err := osOpenFile("/dev/loop-control", osORdOnly, 0644)
f, err := os.OpenFile("/dev/loop-control", os.O_RDONLY, 0644)
if err != nil {
return 0, err
}
@@ -27,27 +30,27 @@ func getNextFreeLoopbackIndex() (int, error) {
return index, err
}
func openNextAvailableLoopback(index int, sparseFile *osFile) (loopFile *osFile, err error) {
func openNextAvailableLoopback(index int, sparseFile *os.File) (loopFile *os.File, err error) {
// Start looking for a free /dev/loop
for {
target := fmt.Sprintf("/dev/loop%d", index)
index++
fi, err := osStat(target)
fi, err := os.Stat(target)
if err != nil {
if osIsNotExist(err) {
if os.IsNotExist(err) {
utils.Errorf("There are no more loopback device available.")
}
return nil, ErrAttachLoopbackDevice
}
if fi.Mode()&osModeDevice != osModeDevice {
if fi.Mode()&os.ModeDevice != os.ModeDevice {
utils.Errorf("Loopback device %s is not a block device.", target)
continue
}
// OpenFile adds O_CLOEXEC
loopFile, err = osOpenFile(target, osORdWr, 0644)
loopFile, err = os.OpenFile(target, os.O_RDWR, 0644)
if err != nil {
utils.Errorf("Error openning loopback device: %s", err)
return nil, ErrAttachLoopbackDevice
@@ -58,7 +61,7 @@ func openNextAvailableLoopback(index int, sparseFile *osFile) (loopFile *osFile,
loopFile.Close()
// If the error is EBUSY, then try the next loopback
if err != sysEBusy {
if err != syscall.EBUSY {
utils.Errorf("Cannot set up loopback device %s: %s", target, err)
return nil, ErrAttachLoopbackDevice
}
@@ -80,8 +83,8 @@ func openNextAvailableLoopback(index int, sparseFile *osFile) (loopFile *osFile,
}
// attachLoopDevice attaches the given sparse file to the next
// available loopback device. It returns an opened *osFile.
func attachLoopDevice(sparseName string) (loop *osFile, err error) {
// available loopback device. It returns an opened *os.File.
func attachLoopDevice(sparseName string) (loop *os.File, err error) {
// Try to retrieve the next available loopback device via syscall.
// If it fails, we discard error and start loopking for a
@@ -92,7 +95,7 @@ func attachLoopDevice(sparseName string) (loop *osFile, err error) {
}
// OpenFile adds O_CLOEXEC
sparseFile, err := osOpenFile(sparseName, osORdWr, 0644)
sparseFile, err := os.OpenFile(sparseName, os.O_RDWR, 0644)
if err != nil {
utils.Errorf("Error openning sparse file %s: %s", sparseName, err)
return nil, ErrAttachLoopbackDevice
+297 -158
View File
@@ -8,6 +8,8 @@ import (
"fmt"
"io"
"io/ioutil"
"os"
"os/exec"
"path"
"path/filepath"
"strconv"
@@ -16,7 +18,9 @@ import (
"syscall"
"time"
"github.com/dotcloud/docker/daemon/graphdriver"
"github.com/dotcloud/docker/pkg/label"
"github.com/dotcloud/docker/pkg/units"
"github.com/dotcloud/docker/utils"
)
@@ -62,8 +66,18 @@ type DeviceSet struct {
devicePrefix string
TransactionId uint64
NewTransactionId uint64
nextFreeDevice int
sawBusy bool
nextDeviceId int
// Options
dataLoopbackSize int64
metaDataLoopbackSize int64
baseFsSize uint64
filesystem string
mountOptions string
mkfsArgs []string
dataDevice string
metadataDevice string
doBlkDiscard bool
}
type DiskUsage struct {
@@ -109,7 +123,19 @@ func (devices *DeviceSet) loopbackDir() string {
return path.Join(devices.root, "devicemapper")
}
func (devices *DeviceSet) jsonFile() string {
func (devices *DeviceSet) metadataDir() string {
return path.Join(devices.root, "metadata")
}
func (devices *DeviceSet) metadataFile(info *DevInfo) string {
file := info.Hash
if file == "" {
file = "base"
}
return path.Join(devices.metadataDir(), file)
}
func (devices *DeviceSet) oldMetadataFile() string {
return path.Join(devices.loopbackDir(), "json")
}
@@ -125,7 +151,7 @@ func (devices *DeviceSet) hasImage(name string) bool {
dirname := devices.loopbackDir()
filename := path.Join(dirname, name)
_, err := osStat(filename)
_, err := os.Stat(filename)
return err == nil
}
@@ -137,16 +163,16 @@ func (devices *DeviceSet) ensureImage(name string, size int64) (string, error) {
dirname := devices.loopbackDir()
filename := path.Join(dirname, name)
if err := osMkdirAll(dirname, 0700); err != nil && !osIsExist(err) {
if err := os.MkdirAll(dirname, 0700); err != nil && !os.IsExist(err) {
return "", err
}
if _, err := osStat(filename); err != nil {
if !osIsNotExist(err) {
if _, err := os.Stat(filename); err != nil {
if !os.IsNotExist(err) {
return "", err
}
utils.Debugf("Creating loopback file %s for device-manage use", filename)
file, err := osOpenFile(filename, osORdWr|osOCreate, 0600)
file, err := os.OpenFile(filename, os.O_RDWR|os.O_CREATE, 0600)
if err != nil {
return "", err
}
@@ -159,26 +185,24 @@ func (devices *DeviceSet) ensureImage(name string, size int64) (string, error) {
return filename, nil
}
func (devices *DeviceSet) allocateDeviceId() int {
// TODO: Add smarter reuse of deleted devices
id := devices.nextFreeDevice
devices.nextFreeDevice = devices.nextFreeDevice + 1
return id
}
func (devices *DeviceSet) allocateTransactionId() uint64 {
devices.NewTransactionId = devices.NewTransactionId + 1
return devices.NewTransactionId
}
func (devices *DeviceSet) saveMetadata() error {
devices.devicesLock.Lock()
jsonData, err := json.Marshal(devices.MetaData)
devices.devicesLock.Unlock()
func (devices *DeviceSet) removeMetadata(info *DevInfo) error {
if err := os.RemoveAll(devices.metadataFile(info)); err != nil {
return fmt.Errorf("Error removing metadata file %s: %s", devices.metadataFile(info), err)
}
return nil
}
func (devices *DeviceSet) saveMetadata(info *DevInfo) error {
jsonData, err := json.Marshal(info)
if err != nil {
return fmt.Errorf("Error encoding metadata to json: %s", err)
}
tmpFile, err := ioutil.TempFile(filepath.Dir(devices.jsonFile()), ".json")
tmpFile, err := ioutil.TempFile(devices.metadataDir(), ".tmp")
if err != nil {
return fmt.Errorf("Error creating metadata file: %s", err)
}
@@ -196,7 +220,7 @@ func (devices *DeviceSet) saveMetadata() error {
if err := tmpFile.Close(); err != nil {
return fmt.Errorf("Error closing metadata file %s: %s", tmpFile.Name(), err)
}
if err := osRename(tmpFile.Name(), devices.jsonFile()); err != nil {
if err := os.Rename(tmpFile.Name(), devices.metadataFile(info)); err != nil {
return fmt.Errorf("Error committing metadata file %s: %s", tmpFile.Name(), err)
}
@@ -214,7 +238,12 @@ func (devices *DeviceSet) lookupDevice(hash string) (*DevInfo, error) {
defer devices.devicesLock.Unlock()
info := devices.Devices[hash]
if info == nil {
return nil, fmt.Errorf("Unknown device %s", hash)
info = devices.loadMetadata(hash)
if info == nil {
return nil, fmt.Errorf("Unknown device %s", hash)
}
devices.Devices[hash] = info
}
return info, nil
}
@@ -234,7 +263,7 @@ func (devices *DeviceSet) registerDevice(id int, hash string, size uint64) (*Dev
devices.Devices[hash] = info
devices.devicesLock.Unlock()
if err := devices.saveMetadata(); err != nil {
if err := devices.saveMetadata(info); err != nil {
// Try to remove unused device
devices.devicesLock.Lock()
delete(devices.Devices, hash)
@@ -258,63 +287,94 @@ func (devices *DeviceSet) activateDeviceIfNeeded(info *DevInfo) error {
func (devices *DeviceSet) createFilesystem(info *DevInfo) error {
devname := info.DevName()
err := execRun("mkfs.ext4", "-E", "discard,lazy_itable_init=0,lazy_journal_init=0", devname)
if err != nil {
err = execRun("mkfs.ext4", "-E", "discard,lazy_itable_init=0", devname)
args := []string{}
for _, arg := range devices.mkfsArgs {
args = append(args, arg)
}
args = append(args, devname)
var err error
switch devices.filesystem {
case "xfs":
err = exec.Command("mkfs.xfs", args...).Run()
case "ext4":
err = exec.Command("mkfs.ext4", append([]string{"-E", "nodiscard,lazy_itable_init=0,lazy_journal_init=0"}, args...)...).Run()
if err != nil {
err = exec.Command("mkfs.ext4", append([]string{"-E", "nodiscard,lazy_itable_init=0"}, args...)...).Run()
}
default:
err = fmt.Errorf("Unsupported filesystem type %s", devices.filesystem)
}
if err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
return nil
}
func (devices *DeviceSet) loadMetaData() error {
utils.Debugf("loadMetadata()")
defer utils.Debugf("loadMetadata END")
func (devices *DeviceSet) initMetaData() error {
_, _, _, params, err := getStatus(devices.getPoolName())
if err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
if _, err := fmt.Sscanf(params, "%d", &devices.TransactionId); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
devices.NewTransactionId = devices.TransactionId
jsonData, err := ioutil.ReadFile(devices.jsonFile())
if err != nil && !osIsNotExist(err) {
utils.Debugf("\n--->Err: %s\n", err)
// Migrate old metadatafile
jsonData, err := ioutil.ReadFile(devices.oldMetadataFile())
if err != nil && !os.IsNotExist(err) {
return err
}
devices.MetaData.Devices = make(map[string]*DevInfo)
if jsonData != nil {
if err := json.Unmarshal(jsonData, &devices.MetaData); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
m := MetaData{Devices: make(map[string]*DevInfo)}
if err := json.Unmarshal(jsonData, &m); err != nil {
return err
}
}
for hash, d := range devices.Devices {
d.Hash = hash
d.devices = devices
for hash, info := range m.Devices {
info.Hash = hash
if d.DeviceId >= devices.nextFreeDevice {
devices.nextFreeDevice = d.DeviceId + 1
// If the transaction id is larger than the actual one we lost the device due to some crash
if info.TransactionId <= devices.TransactionId {
devices.saveMetadata(info)
}
}
if err := os.Rename(devices.oldMetadataFile(), devices.oldMetadataFile()+".migrated"); err != nil {
return err
}
// If the transaction id is larger than the actual one we lost the device due to some crash
if d.TransactionId > devices.TransactionId {
utils.Debugf("Removing lost device %s with id %d", hash, d.TransactionId)
delete(devices.Devices, hash)
}
}
return nil
}
func (devices *DeviceSet) loadMetadata(hash string) *DevInfo {
info := &DevInfo{Hash: hash, devices: devices}
jsonData, err := ioutil.ReadFile(devices.metadataFile(info))
if err != nil {
return nil
}
if err := json.Unmarshal(jsonData, &info); err != nil {
return nil
}
// If the transaction id is larger than the actual one we lost the device due to some crash
if info.TransactionId > devices.TransactionId {
return nil
}
return info
}
func (devices *DeviceSet) setupBaseImage() error {
oldInfo, _ := devices.lookupDevice("")
if oldInfo != nil && oldInfo.Initialized {
@@ -324,45 +384,42 @@ func (devices *DeviceSet) setupBaseImage() error {
if oldInfo != nil && !oldInfo.Initialized {
utils.Debugf("Removing uninitialized base image")
if err := devices.deleteDevice(oldInfo); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
}
utils.Debugf("Initializing base device-manager snapshot")
id := devices.allocateDeviceId()
id := devices.nextDeviceId
// Create initial device
if err := createDevice(devices.getPoolDevName(), id); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
if err := createDevice(devices.getPoolDevName(), &id); err != nil {
return err
}
utils.Debugf("Registering base device (id %v) with FS size %v", id, DefaultBaseFsSize)
info, err := devices.registerDevice(id, "", DefaultBaseFsSize)
// Ids are 24bit, so wrap around
devices.nextDeviceId = (id + 1) & 0xffffff
utils.Debugf("Registering base device (id %v) with FS size %v", id, devices.baseFsSize)
info, err := devices.registerDevice(id, "", devices.baseFsSize)
if err != nil {
_ = deleteDevice(devices.getPoolDevName(), id)
utils.Debugf("\n--->Err: %s\n", err)
return err
}
utils.Debugf("Creating filesystem on base device-manager snapshot")
if err = devices.activateDeviceIfNeeded(info); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
if err := devices.createFilesystem(info); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
info.Initialized = true
if err = devices.saveMetadata(); err != nil {
if err = devices.saveMetadata(info); err != nil {
info.Initialized = false
utils.Debugf("\n--->Err: %s\n", err)
return err
}
@@ -372,11 +429,11 @@ func (devices *DeviceSet) setupBaseImage() error {
func setCloseOnExec(name string) {
if fileInfos, _ := ioutil.ReadDir("/proc/self/fd"); fileInfos != nil {
for _, i := range fileInfos {
link, _ := osReadlink(filepath.Join("/proc/self/fd", i.Name()))
link, _ := os.Readlink(filepath.Join("/proc/self/fd", i.Name()))
if link == name {
fd, err := strconv.Atoi(i.Name())
if err == nil {
sysCloseOnExec(fd)
syscall.CloseOnExec(fd)
}
}
}
@@ -388,10 +445,6 @@ func (devices *DeviceSet) log(level int, file string, line int, dmError int, mes
return // Ignore _LOG_DEBUG
}
if strings.Contains(message, "busy") {
devices.sawBusy = true
}
utils.Debugf("libdevmapper(%d): %s:%d (%d) %s", level, file, line, dmError, message)
}
@@ -408,7 +461,7 @@ func (devices *DeviceSet) ResizePool(size int64) error {
datafilename := path.Join(dirname, "data")
metadatafilename := path.Join(dirname, "metadata")
datafile, err := osOpenFile(datafilename, osORdWr, 0)
datafile, err := os.OpenFile(datafilename, os.O_RDWR, 0)
if datafile == nil {
return err
}
@@ -429,7 +482,7 @@ func (devices *DeviceSet) ResizePool(size int64) error {
}
defer dataloopback.Close()
metadatafile, err := osOpenFile(metadatafilename, osORdWr, 0)
metadatafile, err := os.OpenFile(metadatafilename, os.O_RDWR, 0)
if metadatafile == nil {
return err
}
@@ -472,39 +525,23 @@ func (devices *DeviceSet) ResizePool(size int64) error {
func (devices *DeviceSet) initDevmapper(doInit bool) error {
logInit(devices)
// Make sure the sparse images exist in <root>/devicemapper/data and
// <root>/devicemapper/metadata
hasData := devices.hasImage("data")
hasMetadata := devices.hasImage("metadata")
if !doInit && !hasData {
return errors.New("Loopback data file not found")
}
if !doInit && !hasMetadata {
return errors.New("Loopback metadata file not found")
}
createdLoopback := !hasData || !hasMetadata
data, err := devices.ensureImage("data", DefaultDataLoopbackSize)
_, err := getDriverVersion()
if err != nil {
utils.Debugf("Error device ensureImage (data): %s\n", err)
return err
// Can't even get driver version, assume not supported
return graphdriver.ErrNotSupported
}
metadata, err := devices.ensureImage("metadata", DefaultMetaDataLoopbackSize)
if err != nil {
utils.Debugf("Error device ensureImage (metadata): %s\n", err)
if err := os.MkdirAll(devices.metadataDir(), 0700); err != nil && !os.IsExist(err) {
return err
}
// Set the device prefix from the device id and inode of the docker root dir
st, err := osStat(devices.root)
st, err := os.Stat(devices.root)
if err != nil {
return fmt.Errorf("Error looking up dir %s: %s", devices.root, err)
}
sysSt := toSysStatT(st.Sys())
sysSt := st.Sys().(*syscall.Stat_t)
// "reg-" stands for "regular file".
// In the future we might use "dev-" for "device file", etc.
// docker-maj,min[-inode] stands for:
@@ -527,35 +564,91 @@ func (devices *DeviceSet) initDevmapper(doInit bool) error {
// so we add this badhack to make sure it closes itself
setCloseOnExec("/dev/mapper/control")
// Make sure the sparse images exist in <root>/devicemapper/data and
// <root>/devicemapper/metadata
createdLoopback := false
// If the pool doesn't exist, create it
if info.Exists == 0 {
utils.Debugf("Pool doesn't exist. Creating it.")
dataFile, err := attachLoopDevice(data)
if err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
var (
dataFile *os.File
metadataFile *os.File
)
if devices.dataDevice == "" {
// Make sure the sparse images exist in <root>/devicemapper/data
hasData := devices.hasImage("data")
if !doInit && !hasData {
return errors.New("Loopback data file not found")
}
if !hasData {
createdLoopback = true
}
data, err := devices.ensureImage("data", devices.dataLoopbackSize)
if err != nil {
utils.Debugf("Error device ensureImage (data): %s\n", err)
return err
}
dataFile, err = attachLoopDevice(data)
if err != nil {
return err
}
} else {
dataFile, err = os.OpenFile(devices.dataDevice, os.O_RDWR, 0600)
if err != nil {
return err
}
}
defer dataFile.Close()
metadataFile, err := attachLoopDevice(metadata)
if err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
if devices.metadataDevice == "" {
// Make sure the sparse images exist in <root>/devicemapper/metadata
hasMetadata := devices.hasImage("metadata")
if !doInit && !hasMetadata {
return errors.New("Loopback metadata file not found")
}
if !hasMetadata {
createdLoopback = true
}
metadata, err := devices.ensureImage("metadata", devices.metaDataLoopbackSize)
if err != nil {
utils.Debugf("Error device ensureImage (metadata): %s\n", err)
return err
}
metadataFile, err = attachLoopDevice(metadata)
if err != nil {
return err
}
} else {
metadataFile, err = os.OpenFile(devices.metadataDevice, os.O_RDWR, 0600)
if err != nil {
return err
}
}
defer metadataFile.Close()
if err := createPool(devices.getPoolName(), dataFile, metadataFile); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
}
// If we didn't just create the data or metadata image, we need to
// load the metadata from the existing file.
// load the transaction id and migrate old metadata
if !createdLoopback {
if err = devices.loadMetaData(); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
if err = devices.initMetaData(); err != nil {
return err
}
}
@@ -587,13 +680,16 @@ func (devices *DeviceSet) AddDevice(hash, baseHash string) error {
return fmt.Errorf("device %s already exists", hash)
}
deviceId := devices.allocateDeviceId()
deviceId := devices.nextDeviceId
if err := devices.createSnapDevice(devices.getPoolDevName(), deviceId, baseInfo.Name(), baseInfo.DeviceId); err != nil {
if err := createSnapDevice(devices.getPoolDevName(), &deviceId, baseInfo.Name(), baseInfo.DeviceId); err != nil {
utils.Debugf("Error creating snap device: %s\n", err)
return err
}
// Ids are 24bit, so wrap around
devices.nextDeviceId = (deviceId + 1) & 0xffffff
if _, err := devices.registerDevice(deviceId, hash, baseInfo.Size); err != nil {
deleteDevice(devices.getPoolDevName(), deviceId)
utils.Debugf("Error registering device: %s\n", err)
@@ -603,12 +699,14 @@ func (devices *DeviceSet) AddDevice(hash, baseHash string) error {
}
func (devices *DeviceSet) deleteDevice(info *DevInfo) error {
// This is a workaround for the kernel not discarding block so
// on the thin pool when we remove a thinp device, so we do it
// manually
if err := devices.activateDeviceIfNeeded(info); err == nil {
if err := BlockDeviceDiscard(info.DevName()); err != nil {
utils.Debugf("Error discarding block on device: %s (ignoring)\n", err)
if devices.doBlkDiscard {
// This is a workaround for the kernel not discarding block so
// on the thin pool when we remove a thinp device, so we do it
// manually
if err := devices.activateDeviceIfNeeded(info); err == nil {
if err := BlockDeviceDiscard(info.DevName()); err != nil {
utils.Debugf("Error discarding block on device: %s (ignoring)\n", err)
}
}
}
@@ -620,14 +718,6 @@ func (devices *DeviceSet) deleteDevice(info *DevInfo) error {
}
}
if info.Initialized {
info.Initialized = false
if err := devices.saveMetadata(); err != nil {
utils.Debugf("Error saving meta data: %s\n", err)
return err
}
}
if err := deleteDevice(devices.getPoolDevName(), info.DeviceId); err != nil {
utils.Debugf("Error deleting device: %s\n", err)
return err
@@ -638,11 +728,11 @@ func (devices *DeviceSet) deleteDevice(info *DevInfo) error {
delete(devices.Devices, info.Hash)
devices.devicesLock.Unlock()
if err := devices.saveMetadata(); err != nil {
if err := devices.removeMetadata(info); err != nil {
devices.devicesLock.Lock()
devices.Devices[info.Hash] = info
devices.devicesLock.Unlock()
utils.Debugf("Error saving meta data: %s\n", err)
utils.Debugf("Error removing meta data: %s\n", err)
return err
}
@@ -670,7 +760,6 @@ func (devices *DeviceSet) deactivatePool() error {
devname := devices.getPoolDevName()
devinfo, err := getInfo(devname)
if err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
if devinfo.Exists != 0 {
@@ -692,12 +781,10 @@ func (devices *DeviceSet) deactivateDevice(info *DevInfo) error {
devinfo, err := getInfo(info.Name())
if err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
if devinfo.Exists != 0 {
if err := devices.removeDeviceAndWait(info.Name()); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
return err
}
}
@@ -711,12 +798,11 @@ func (devices *DeviceSet) removeDeviceAndWait(devname string) error {
var err error
for i := 0; i < 1000; i++ {
devices.sawBusy = false
err = removeDevice(devname)
if err == nil {
break
}
if !devices.sawBusy {
if err != ErrBusy {
return err
}
@@ -813,7 +899,7 @@ func (devices *DeviceSet) Shutdown() error {
// We use MNT_DETACH here in case it is still busy in some running
// container. This means it'll go away from the global scope directly,
// and the device will be released when that container dies.
if err := sysUnmount(info.mountPath, syscall.MNT_DETACH); err != nil {
if err := syscall.Unmount(info.mountPath, syscall.MNT_DETACH); err != nil {
utils.Debugf("Shutdown unmounting %s, error: %s\n", info.mountPath, err)
}
@@ -871,13 +957,26 @@ func (devices *DeviceSet) MountDevice(hash, path, mountLabel string) error {
return fmt.Errorf("Error activating devmapper device for '%s': %s", hash, err)
}
var flags uintptr = sysMsMgcVal
var flags uintptr = syscall.MS_MGC_VAL
mountOptions := label.FormatMountLabel("discard", mountLabel)
err = sysMount(info.DevName(), path, "ext4", flags, mountOptions)
if err != nil && err == sysEInval {
mountOptions = label.FormatMountLabel("", mountLabel)
err = sysMount(info.DevName(), path, "ext4", flags, mountOptions)
fstype, err := ProbeFsType(info.DevName())
if err != nil {
return err
}
options := ""
if fstype == "xfs" {
// XFS needs nouuid or it can't mount filesystems with the same fs
options = joinMountOptions(options, "nouuid")
}
options = joinMountOptions(options, devices.mountOptions)
options = joinMountOptions(options, label.FormatMountLabel("", mountLabel))
err = syscall.Mount(info.DevName(), path, fstype, flags, joinMountOptions("discard", options))
if err != nil && err == syscall.EINVAL {
err = syscall.Mount(info.DevName(), path, fstype, flags, options)
}
if err != nil {
return fmt.Errorf("Error mounting '%s' on '%s': %s", info.DevName(), path, err)
@@ -886,7 +985,7 @@ func (devices *DeviceSet) MountDevice(hash, path, mountLabel string) error {
info.mountCount = 1
info.mountPath = path
return devices.setInitialized(info)
return nil
}
func (devices *DeviceSet) UnmountDevice(hash string) error {
@@ -914,8 +1013,7 @@ func (devices *DeviceSet) UnmountDevice(hash string) error {
}
utils.Debugf("[devmapper] Unmount(%s)", info.mountPath)
if err := sysUnmount(info.mountPath, 0); err != nil {
utils.Debugf("\n--->Err: %s\n", err)
if err := syscall.Unmount(info.mountPath, 0); err != nil {
return err
}
utils.Debugf("[devmapper] Unmount done")
@@ -937,14 +1035,6 @@ func (devices *DeviceSet) HasDevice(hash string) bool {
return info != nil
}
func (devices *DeviceSet) HasInitializedDevice(hash string) bool {
devices.Lock()
defer devices.Unlock()
info, _ := devices.lookupDevice(hash)
return info != nil && info.Initialized
}
func (devices *DeviceSet) HasActivatedDevice(hash string) bool {
info, _ := devices.lookupDevice(hash)
if info == nil {
@@ -961,17 +1051,6 @@ func (devices *DeviceSet) HasActivatedDevice(hash string) bool {
return devinfo != nil && devinfo.Exists != 0
}
func (devices *DeviceSet) setInitialized(info *DevInfo) error {
info.Initialized = true
if err := devices.saveMetadata(); err != nil {
info.Initialized = false
utils.Debugf("\n--->Err: %s\n", err)
return err
}
return nil
}
func (devices *DeviceSet) List() []string {
devices.Lock()
defer devices.Unlock()
@@ -1069,12 +1148,72 @@ func (devices *DeviceSet) Status() *Status {
return status
}
func NewDeviceSet(root string, doInit bool) (*DeviceSet, error) {
func NewDeviceSet(root string, doInit bool, options []string) (*DeviceSet, error) {
SetDevDir("/dev")
devices := &DeviceSet{
root: root,
MetaData: MetaData{Devices: make(map[string]*DevInfo)},
root: root,
MetaData: MetaData{Devices: make(map[string]*DevInfo)},
dataLoopbackSize: DefaultDataLoopbackSize,
metaDataLoopbackSize: DefaultMetaDataLoopbackSize,
baseFsSize: DefaultBaseFsSize,
filesystem: "ext4",
doBlkDiscard: true,
}
foundBlkDiscard := false
for _, option := range options {
key, val, err := utils.ParseKeyValueOpt(option)
if err != nil {
return nil, err
}
key = strings.ToLower(key)
switch key {
case "dm.basesize":
size, err := units.FromHumanSize(val)
if err != nil {
return nil, err
}
devices.baseFsSize = uint64(size)
case "dm.loopdatasize":
size, err := units.FromHumanSize(val)
if err != nil {
return nil, err
}
devices.dataLoopbackSize = size
case "dm.loopmetadatasize":
size, err := units.FromHumanSize(val)
if err != nil {
return nil, err
}
devices.metaDataLoopbackSize = size
case "dm.fs":
if val != "ext4" && val != "xfs" {
return nil, fmt.Errorf("Unsupported filesystem %s\n", val)
}
devices.filesystem = val
case "dm.mkfsarg":
devices.mkfsArgs = append(devices.mkfsArgs, val)
case "dm.mountopt":
devices.mountOptions = joinMountOptions(devices.mountOptions, val)
case "dm.metadatadev":
devices.metadataDevice = val
case "dm.datadev":
devices.dataDevice = val
case "dm.blkdiscard":
foundBlkDiscard = true
devices.doBlkDiscard, err = strconv.ParseBool(val)
if err != nil {
return nil, err
}
default:
return nil, fmt.Errorf("Unknown option %s\n", key)
}
}
// By default, don't do blk discard hack on raw devices, its rarely useful and is expensive
if !foundBlkDiscard && devices.dataDevice != "" {
devices.doBlkDiscard = false
}
if err := devices.initDevmapper(doInit); err != nil {
+119 -68
View File
@@ -5,9 +5,11 @@ package devmapper
import (
"errors"
"fmt"
"github.com/dotcloud/docker/utils"
"os"
"runtime"
"syscall"
"github.com/dotcloud/docker/utils"
)
type DevmapperLogger interface {
@@ -50,6 +52,7 @@ var (
ErrTaskAddTarget = errors.New("dm_task_add_target failed")
ErrTaskSetSector = errors.New("dm_task_set_sector failed")
ErrTaskGetInfo = errors.New("dm_task_get_info failed")
ErrTaskGetDriverVersion = errors.New("dm_task_get_driver_version failed")
ErrTaskSetCookie = errors.New("dm_task_set_cookie failed")
ErrNilCookie = errors.New("cookie ptr can't be nil")
ErrAttachLoopbackDevice = errors.New("loopback mounting failed")
@@ -62,6 +65,10 @@ var (
ErrInvalidAddNode = errors.New("Invalide AddNoce type")
ErrGetLoopbackBackingFile = errors.New("Unable to get loopback backing file")
ErrLoopbackSetCapacity = errors.New("Unable set loopback capacity")
ErrBusy = errors.New("Device is Busy")
dmSawBusy bool
dmSawExist bool
)
type (
@@ -172,6 +179,14 @@ func (t *Task) GetInfo() (*Info, error) {
return info, nil
}
func (t *Task) GetDriverVersion() (string, error) {
res := DmTaskGetDriverVersion(t.unmanaged)
if res == "" {
return "", ErrTaskGetDriverVersion
}
return res, nil
}
func (t *Task) GetNextTarget(next uintptr) (nextPtr uintptr, start uint64,
length uint64, targetType string, params string) {
@@ -180,7 +195,7 @@ func (t *Task) GetNextTarget(next uintptr) (nextPtr uintptr, start uint64,
start, length, targetType, params
}
func getLoopbackBackingFile(file *osFile) (uint64, uint64, error) {
func getLoopbackBackingFile(file *os.File) (uint64, uint64, error) {
loopInfo, err := ioctlLoopGetStatus64(file.Fd())
if err != nil {
utils.Errorf("Error get loopback backing file: %s\n", err)
@@ -189,7 +204,7 @@ func getLoopbackBackingFile(file *osFile) (uint64, uint64, error) {
return loopInfo.loDevice, loopInfo.loInode, nil
}
func LoopbackSetCapacity(file *osFile) error {
func LoopbackSetCapacity(file *os.File) error {
if err := ioctlLoopSetCapacity(file.Fd(), 0); err != nil {
utils.Errorf("Error loopbackSetCapacity: %s", err)
return ErrLoopbackSetCapacity
@@ -197,20 +212,20 @@ func LoopbackSetCapacity(file *osFile) error {
return nil
}
func FindLoopDeviceFor(file *osFile) *osFile {
func FindLoopDeviceFor(file *os.File) *os.File {
stat, err := file.Stat()
if err != nil {
return nil
}
targetInode := stat.Sys().(*sysStatT).Ino
targetDevice := stat.Sys().(*sysStatT).Dev
targetInode := stat.Sys().(*syscall.Stat_t).Ino
targetDevice := stat.Sys().(*syscall.Stat_t).Dev
for i := 0; true; i++ {
path := fmt.Sprintf("/dev/loop%d", i)
file, err := osOpenFile(path, osORdWr, 0)
file, err := os.OpenFile(path, os.O_RDWR, 0)
if err != nil {
if osIsNotExist(err) {
if os.IsNotExist(err) {
return nil
}
@@ -280,7 +295,7 @@ func RemoveDevice(name string) error {
return nil
}
func GetBlockDeviceSize(file *osFile) (uint64, error) {
func GetBlockDeviceSize(file *os.File) (uint64, error) {
size, err := ioctlBlkGetSize64(file.Fd())
if err != nil {
utils.Errorf("Error getblockdevicesize: %s", err)
@@ -290,7 +305,7 @@ func GetBlockDeviceSize(file *osFile) (uint64, error) {
}
func BlockDeviceDiscard(path string) error {
file, err := osOpenFile(path, osORdWr, 0)
file, err := os.OpenFile(path, os.O_RDWR, 0)
if err != nil {
return err
}
@@ -313,7 +328,7 @@ func BlockDeviceDiscard(path string) error {
}
// This is the programmatic example of "dmsetup create"
func createPool(poolName string, dataFile, metadataFile *osFile) error {
func createPool(poolName string, dataFile, metadataFile *os.File) error {
task, err := createTask(DeviceCreate, poolName)
if task == nil {
return err
@@ -321,21 +336,21 @@ func createPool(poolName string, dataFile, metadataFile *osFile) error {
size, err := GetBlockDeviceSize(dataFile)
if err != nil {
return fmt.Errorf("Can't get data size")
return fmt.Errorf("Can't get data size %s", err)
}
params := metadataFile.Name() + " " + dataFile.Name() + " 128 32768 1 skip_block_zeroing"
if err := task.AddTarget(0, size/512, "thin-pool", params); err != nil {
return fmt.Errorf("Can't add target")
return fmt.Errorf("Can't add target %s", err)
}
var cookie uint = 0
if err := task.SetCookie(&cookie, 0); err != nil {
return fmt.Errorf("Can't set cookie")
return fmt.Errorf("Can't set cookie %s", err)
}
if err := task.Run(); err != nil {
return fmt.Errorf("Error running DeviceCreate (createPool)")
return fmt.Errorf("Error running DeviceCreate (createPool) %s", err)
}
UdevWait(cookie)
@@ -343,7 +358,7 @@ func createPool(poolName string, dataFile, metadataFile *osFile) error {
return nil
}
func reloadPool(poolName string, dataFile, metadataFile *osFile) error {
func reloadPool(poolName string, dataFile, metadataFile *os.File) error {
task, err := createTask(DeviceReload, poolName)
if task == nil {
return err
@@ -351,16 +366,16 @@ func reloadPool(poolName string, dataFile, metadataFile *osFile) error {
size, err := GetBlockDeviceSize(dataFile)
if err != nil {
return fmt.Errorf("Can't get data size")
return fmt.Errorf("Can't get data size %s", err)
}
params := metadataFile.Name() + " " + dataFile.Name() + " 128 32768"
if err := task.AddTarget(0, size/512, "thin-pool", params); err != nil {
return fmt.Errorf("Can't add target")
return fmt.Errorf("Can't add target %s", err)
}
if err := task.Run(); err != nil {
return fmt.Errorf("Error running DeviceCreate")
return fmt.Errorf("Error running DeviceCreate %s", err)
}
return nil
@@ -388,6 +403,17 @@ func getInfo(name string) (*Info, error) {
return task.GetInfo()
}
func getDriverVersion() (string, error) {
task := TaskCreate(DeviceVersion)
if task == nil {
return "", fmt.Errorf("Can't create DeviceVersion task")
}
if err := task.Run(); err != nil {
return "", err
}
return task.GetDriverVersion()
}
func getStatus(name string) (uint64, uint64, string, string, error) {
task, err := createTask(DeviceStatus, name)
if task == nil {
@@ -420,15 +446,15 @@ func setTransactionId(poolName string, oldId uint64, newId uint64) error {
}
if err := task.SetSector(0); err != nil {
return fmt.Errorf("Can't set sector")
return fmt.Errorf("Can't set sector %s", err)
}
if err := task.SetMessage(fmt.Sprintf("set_transaction_id %d %d", oldId, newId)); err != nil {
return fmt.Errorf("Can't set message")
return fmt.Errorf("Can't set message %s", err)
}
if err := task.Run(); err != nil {
return fmt.Errorf("Error running setTransactionId")
return fmt.Errorf("Error running setTransactionId %s", err)
}
return nil
}
@@ -439,7 +465,7 @@ func suspendDevice(name string) error {
return err
}
if err := task.Run(); err != nil {
return fmt.Errorf("Error running DeviceSuspend: %s", err)
return fmt.Errorf("Error running DeviceSuspend %s", err)
}
return nil
}
@@ -452,11 +478,11 @@ func resumeDevice(name string) error {
var cookie uint = 0
if err := task.SetCookie(&cookie, 0); err != nil {
return fmt.Errorf("Can't set cookie")
return fmt.Errorf("Can't set cookie %s", err)
}
if err := task.Run(); err != nil {
return fmt.Errorf("Error running DeviceResume")
return fmt.Errorf("Error running DeviceResume %s", err)
}
UdevWait(cookie)
@@ -464,23 +490,33 @@ func resumeDevice(name string) error {
return nil
}
func createDevice(poolName string, deviceId int) error {
utils.Debugf("[devmapper] createDevice(poolName=%v, deviceId=%v)", poolName, deviceId)
task, err := createTask(DeviceTargetMsg, poolName)
if task == nil {
return err
}
func createDevice(poolName string, deviceId *int) error {
utils.Debugf("[devmapper] createDevice(poolName=%v, deviceId=%v)", poolName, *deviceId)
if err := task.SetSector(0); err != nil {
return fmt.Errorf("Can't set sector")
}
for {
task, err := createTask(DeviceTargetMsg, poolName)
if task == nil {
return err
}
if err := task.SetMessage(fmt.Sprintf("create_thin %d", deviceId)); err != nil {
return fmt.Errorf("Can't set message")
}
if err := task.SetSector(0); err != nil {
return fmt.Errorf("Can't set sector %s", err)
}
if err := task.Run(); err != nil {
return fmt.Errorf("Error running createDevice")
if err := task.SetMessage(fmt.Sprintf("create_thin %d", *deviceId)); err != nil {
return fmt.Errorf("Can't set message %s", err)
}
dmSawExist = false
if err := task.Run(); err != nil {
if dmSawExist {
// Already exists, try next id
*deviceId++
continue
}
return fmt.Errorf("Error running createDevice %s", err)
}
break
}
return nil
}
@@ -492,15 +528,15 @@ func deleteDevice(poolName string, deviceId int) error {
}
if err := task.SetSector(0); err != nil {
return fmt.Errorf("Can't set sector")
return fmt.Errorf("Can't set sector %s", err)
}
if err := task.SetMessage(fmt.Sprintf("delete %d", deviceId)); err != nil {
return fmt.Errorf("Can't set message")
return fmt.Errorf("Can't set message %s", err)
}
if err := task.Run(); err != nil {
return fmt.Errorf("Error running deleteDevice")
return fmt.Errorf("Error running deleteDevice %s", err)
}
return nil
}
@@ -512,8 +548,12 @@ func removeDevice(name string) error {
if task == nil {
return err
}
dmSawBusy = false
if err = task.Run(); err != nil {
return fmt.Errorf("Error running removeDevice")
if dmSawBusy {
return ErrBusy
}
return fmt.Errorf("Error running removeDevice %s", err)
}
return nil
}
@@ -526,19 +566,19 @@ func activateDevice(poolName string, name string, deviceId int, size uint64) err
params := fmt.Sprintf("%s %d", poolName, deviceId)
if err := task.AddTarget(0, size/512, "thin", params); err != nil {
return fmt.Errorf("Can't add target")
return fmt.Errorf("Can't add target %s", err)
}
if err := task.SetAddNode(AddNodeOnCreate); err != nil {
return fmt.Errorf("Can't add node")
return fmt.Errorf("Can't add node %s", err)
}
var cookie uint = 0
if err := task.SetCookie(&cookie, 0); err != nil {
return fmt.Errorf("Can't set cookie")
return fmt.Errorf("Can't set cookie %s", err)
}
if err := task.Run(); err != nil {
return fmt.Errorf("Error running DeviceCreate (activateDevice)")
return fmt.Errorf("Error running DeviceCreate (activateDevice) %s", err)
}
UdevWait(cookie)
@@ -546,7 +586,7 @@ func activateDevice(poolName string, name string, deviceId int, size uint64) err
return nil
}
func (devices *DeviceSet) createSnapDevice(poolName string, deviceId int, baseName string, baseDeviceId int) error {
func createSnapDevice(poolName string, deviceId *int, baseName string, baseDeviceId int) error {
devinfo, _ := getInfo(baseName)
doSuspend := devinfo != nil && devinfo.Exists != 0
@@ -556,33 +596,44 @@ func (devices *DeviceSet) createSnapDevice(poolName string, deviceId int, baseNa
}
}
task, err := createTask(DeviceTargetMsg, poolName)
if task == nil {
if doSuspend {
resumeDevice(baseName)
for {
task, err := createTask(DeviceTargetMsg, poolName)
if task == nil {
if doSuspend {
resumeDevice(baseName)
}
return err
}
return err
}
if err := task.SetSector(0); err != nil {
if doSuspend {
resumeDevice(baseName)
if err := task.SetSector(0); err != nil {
if doSuspend {
resumeDevice(baseName)
}
return fmt.Errorf("Can't set sector %s", err)
}
return fmt.Errorf("Can't set sector")
}
if err := task.SetMessage(fmt.Sprintf("create_snap %d %d", deviceId, baseDeviceId)); err != nil {
if doSuspend {
resumeDevice(baseName)
if err := task.SetMessage(fmt.Sprintf("create_snap %d %d", *deviceId, baseDeviceId)); err != nil {
if doSuspend {
resumeDevice(baseName)
}
return fmt.Errorf("Can't set message %s", err)
}
return fmt.Errorf("Can't set message")
}
if err := task.Run(); err != nil {
if doSuspend {
resumeDevice(baseName)
dmSawExist = false
if err := task.Run(); err != nil {
if dmSawExist {
// Already exists, try next id
*deviceId++
continue
}
if doSuspend {
resumeDevice(baseName)
}
return fmt.Errorf("Error running DeviceCreate (createSnapDevice) %s", err)
}
return fmt.Errorf("Error running DeviceCreate (createSnapDevice)")
break
}
if doSuspend {
+16 -1
View File
@@ -4,12 +4,27 @@ package devmapper
import "C"
import (
"strings"
)
// Due to the way cgo works this has to be in a separate file, as devmapper.go has
// definitions in the cgo block, which is incompatible with using "//export"
//export DevmapperLogCallback
func DevmapperLogCallback(level C.int, file *C.char, line C.int, dm_errno_or_class C.int, message *C.char) {
msg := C.GoString(message)
if level < 7 {
if strings.Contains(msg, "busy") {
dmSawBusy = true
}
if strings.Contains(msg, "File exists") {
dmSawExist = true
}
}
if dmLogger != nil {
dmLogger.log(int(level), C.GoString(file), int(line), int(dm_errno_or_class), C.GoString(message))
dmLogger.log(int(level), C.GoString(file), int(line), int(dm_errno_or_class), msg)
}
}
+18 -268
View File
@@ -3,285 +3,35 @@
package devmapper
import (
"github.com/dotcloud/docker/daemon/graphdriver/graphtest"
"testing"
)
func TestTaskCreate(t *testing.T) {
t.Skip("FIXME: not a unit test")
// Test success
taskCreate(t, DeviceInfo)
// Test Failure
DmTaskCreate = dmTaskCreateFail
defer func() { DmTaskCreate = dmTaskCreateFct }()
if task := TaskCreate(-1); task != nil {
t.Fatalf("An error should have occured while creating an invalid task.")
}
func init() {
// Reduce the size the the base fs and loopback for the tests
DefaultDataLoopbackSize = 300 * 1024 * 1024
DefaultMetaDataLoopbackSize = 200 * 1024 * 1024
DefaultBaseFsSize = 300 * 1024 * 1024
}
func TestTaskRun(t *testing.T) {
t.Skip("FIXME: not a unit test")
task := taskCreate(t, DeviceInfo)
// Test success
// Perform the RUN
if err := task.Run(); err != nil {
t.Fatal(err)
}
// Make sure we don't have error with GetInfo
if _, err := task.GetInfo(); err != nil {
t.Fatal(err)
}
// Test failure
DmTaskRun = dmTaskRunFail
defer func() { DmTaskRun = dmTaskRunFct }()
task = taskCreate(t, DeviceInfo)
// Perform the RUN
if err := task.Run(); err != ErrTaskRun {
t.Fatalf("An error should have occured while running task.")
}
// Make sure GetInfo also fails
if _, err := task.GetInfo(); err != ErrTaskGetInfo {
t.Fatalf("GetInfo should fail if task.Run() failed.")
}
// This avoids creating a new driver for each test if all tests are run
// Make sure to put new tests between TestDevmapperSetup and TestDevmapperTeardown
func TestDevmapperSetup(t *testing.T) {
graphtest.GetDriver(t, "devicemapper")
}
func TestTaskSetName(t *testing.T) {
t.Skip("FIXME: not a unit test")
task := taskCreate(t, DeviceInfo)
// Test success
if err := task.SetName("test"); err != nil {
t.Fatal(err)
}
// Test failure
DmTaskSetName = dmTaskSetNameFail
defer func() { DmTaskSetName = dmTaskSetNameFct }()
if err := task.SetName("test"); err != ErrTaskSetName {
t.Fatalf("An error should have occured while runnign SetName.")
}
func TestDevmapperCreateEmpty(t *testing.T) {
graphtest.DriverTestCreateEmpty(t, "devicemapper")
}
func TestTaskSetMessage(t *testing.T) {
t.Skip("FIXME: not a unit test")
task := taskCreate(t, DeviceInfo)
// Test success
if err := task.SetMessage("test"); err != nil {
t.Fatal(err)
}
// Test failure
DmTaskSetMessage = dmTaskSetMessageFail
defer func() { DmTaskSetMessage = dmTaskSetMessageFct }()
if err := task.SetMessage("test"); err != ErrTaskSetMessage {
t.Fatalf("An error should have occured while runnign SetMessage.")
}
func TestDevmapperCreateBase(t *testing.T) {
graphtest.DriverTestCreateBase(t, "devicemapper")
}
func TestTaskSetSector(t *testing.T) {
t.Skip("FIXME: not a unit test")
task := taskCreate(t, DeviceInfo)
// Test success
if err := task.SetSector(128); err != nil {
t.Fatal(err)
}
DmTaskSetSector = dmTaskSetSectorFail
defer func() { DmTaskSetSector = dmTaskSetSectorFct }()
// Test failure
if err := task.SetSector(0); err != ErrTaskSetSector {
t.Fatalf("An error should have occured while running SetSector.")
}
func TestDevmapperCreateSnap(t *testing.T) {
graphtest.DriverTestCreateSnap(t, "devicemapper")
}
func TestTaskSetCookie(t *testing.T) {
t.Skip("FIXME: not a unit test")
var (
cookie uint = 0
task = taskCreate(t, DeviceInfo)
)
// Test success
if err := task.SetCookie(&cookie, 0); err != nil {
t.Fatal(err)
}
// Test failure
if err := task.SetCookie(nil, 0); err != ErrNilCookie {
t.Fatalf("An error should have occured while running SetCookie with nil cookie.")
}
DmTaskSetCookie = dmTaskSetCookieFail
defer func() { DmTaskSetCookie = dmTaskSetCookieFct }()
if err := task.SetCookie(&cookie, 0); err != ErrTaskSetCookie {
t.Fatalf("An error should have occured while running SetCookie.")
}
}
func TestTaskSetAddNode(t *testing.T) {
t.Skip("FIXME: not a unit test")
task := taskCreate(t, DeviceInfo)
// Test success
if err := task.SetAddNode(0); err != nil {
t.Fatal(err)
}
// Test failure
if err := task.SetAddNode(-1); err != ErrInvalidAddNode {
t.Fatalf("An error should have occured running SetAddNode with wrong node.")
}
DmTaskSetAddNode = dmTaskSetAddNodeFail
defer func() { DmTaskSetAddNode = dmTaskSetAddNodeFct }()
if err := task.SetAddNode(0); err != ErrTaskSetAddNode {
t.Fatalf("An error should have occured running SetAddNode.")
}
}
func TestTaskSetRo(t *testing.T) {
t.Skip("FIXME: not a unit test")
task := taskCreate(t, DeviceInfo)
// Test success
if err := task.SetRo(); err != nil {
t.Fatal(err)
}
// Test failure
DmTaskSetRo = dmTaskSetRoFail
defer func() { DmTaskSetRo = dmTaskSetRoFct }()
if err := task.SetRo(); err != ErrTaskSetRo {
t.Fatalf("An error should have occured running SetRo.")
}
}
func TestTaskAddTarget(t *testing.T) {
t.Skip("FIXME: not a unit test")
task := taskCreate(t, DeviceInfo)
// Test success
if err := task.AddTarget(0, 128, "thinp", ""); err != nil {
t.Fatal(err)
}
// Test failure
DmTaskAddTarget = dmTaskAddTargetFail
defer func() { DmTaskAddTarget = dmTaskAddTargetFct }()
if err := task.AddTarget(0, 128, "thinp", ""); err != ErrTaskAddTarget {
t.Fatalf("An error should have occured running AddTarget.")
}
}
// func TestTaskGetInfo(t *testing.T) {
// task := taskCreate(t, DeviceInfo)
// // Test success
// if _, err := task.GetInfo(); err != nil {
// t.Fatal(err)
// }
// // Test failure
// DmTaskGetInfo = dmTaskGetInfoFail
// defer func() { DmTaskGetInfo = dmTaskGetInfoFct }()
// if _, err := task.GetInfo(); err != ErrTaskGetInfo {
// t.Fatalf("An error should have occured running GetInfo.")
// }
// }
// func TestTaskGetNextTarget(t *testing.T) {
// task := taskCreate(t, DeviceInfo)
// if next, _, _, _, _ := task.GetNextTarget(0); next == 0 {
// t.Fatalf("The next target should not be 0.")
// }
// }
/// Utils
func taskCreate(t *testing.T, taskType TaskType) *Task {
task := TaskCreate(taskType)
if task == nil {
t.Fatalf("Error creating task")
}
return task
}
/// Failure function replacement
func dmTaskCreateFail(t int) *CDmTask {
return nil
}
func dmTaskRunFail(task *CDmTask) int {
return -1
}
func dmTaskSetNameFail(task *CDmTask, name string) int {
return -1
}
func dmTaskSetMessageFail(task *CDmTask, message string) int {
return -1
}
func dmTaskSetSectorFail(task *CDmTask, sector uint64) int {
return -1
}
func dmTaskSetCookieFail(task *CDmTask, cookie *uint, flags uint16) int {
return -1
}
func dmTaskSetAddNodeFail(task *CDmTask, addNode AddNodeType) int {
return -1
}
func dmTaskSetRoFail(task *CDmTask) int {
return -1
}
func dmTaskAddTargetFail(task *CDmTask,
start, size uint64, ttype, params string) int {
return -1
}
func dmTaskGetInfoFail(task *CDmTask, info *Info) int {
return -1
}
func dmGetNextTargetFail(task *CDmTask, next uintptr, start, length *uint64,
target, params *string) uintptr {
return 0
}
func dmAttachLoopDeviceFail(filename string, fd *int) string {
return ""
}
func sysGetBlockSizeFail(fd uintptr, size *uint64) sysErrno {
return 1
}
func dmUdevWaitFail(cookie uint) int {
return -1
}
func dmSetDevDirFail(dir string) int {
return -1
}
func dmGetLibraryVersionFail(version *string) int {
return -1
func TestDevmapperTeardown(t *testing.T) {
graphtest.PutDriver(t)
}
@@ -85,23 +85,24 @@ const (
)
var (
DmGetLibraryVersion = dmGetLibraryVersionFct
DmGetNextTarget = dmGetNextTargetFct
DmLogInitVerbose = dmLogInitVerboseFct
DmSetDevDir = dmSetDevDirFct
DmTaskAddTarget = dmTaskAddTargetFct
DmTaskCreate = dmTaskCreateFct
DmTaskDestroy = dmTaskDestroyFct
DmTaskGetInfo = dmTaskGetInfoFct
DmTaskRun = dmTaskRunFct
DmTaskSetAddNode = dmTaskSetAddNodeFct
DmTaskSetCookie = dmTaskSetCookieFct
DmTaskSetMessage = dmTaskSetMessageFct
DmTaskSetName = dmTaskSetNameFct
DmTaskSetRo = dmTaskSetRoFct
DmTaskSetSector = dmTaskSetSectorFct
DmUdevWait = dmUdevWaitFct
LogWithErrnoInit = logWithErrnoInitFct
DmGetLibraryVersion = dmGetLibraryVersionFct
DmGetNextTarget = dmGetNextTargetFct
DmLogInitVerbose = dmLogInitVerboseFct
DmSetDevDir = dmSetDevDirFct
DmTaskAddTarget = dmTaskAddTargetFct
DmTaskCreate = dmTaskCreateFct
DmTaskDestroy = dmTaskDestroyFct
DmTaskGetInfo = dmTaskGetInfoFct
DmTaskGetDriverVersion = dmTaskGetDriverVersionFct
DmTaskRun = dmTaskRunFct
DmTaskSetAddNode = dmTaskSetAddNodeFct
DmTaskSetCookie = dmTaskSetCookieFct
DmTaskSetMessage = dmTaskSetMessageFct
DmTaskSetName = dmTaskSetNameFct
DmTaskSetRo = dmTaskSetRoFct
DmTaskSetSector = dmTaskSetSectorFct
DmUdevWait = dmUdevWaitFct
LogWithErrnoInit = logWithErrnoInitFct
)
func free(p *C.char) {
@@ -184,6 +185,16 @@ func dmTaskGetInfoFct(task *CDmTask, info *Info) int {
return int(C.dm_task_get_info((*C.struct_dm_task)(task), &Cinfo))
}
func dmTaskGetDriverVersionFct(task *CDmTask) string {
buffer := C.malloc(128)
defer C.free(buffer)
res := C.dm_task_get_driver_version((*C.struct_dm_task)(task), (*C.char)(buffer), 128)
if res == 0 {
return ""
}
return C.GoString((*C.char)(buffer))
}
func dmGetNextTargetFct(task *CDmTask, next uintptr, start, length *uint64, target, params *string) uintptr {
var (
Cstart, Clength C.uint64_t
+19 -6
View File
@@ -9,6 +9,7 @@ import (
"path"
"github.com/dotcloud/docker/daemon/graphdriver"
"github.com/dotcloud/docker/pkg/mount"
"github.com/dotcloud/docker/utils"
)
@@ -26,15 +27,21 @@ type Driver struct {
home string
}
var Init = func(home string) (graphdriver.Driver, error) {
deviceSet, err := NewDeviceSet(home, true)
func Init(home string, options []string) (graphdriver.Driver, error) {
deviceSet, err := NewDeviceSet(home, true, options)
if err != nil {
return nil, err
}
if err := graphdriver.MakePrivate(home); err != nil {
return nil, err
}
d := &Driver{
DeviceSet: deviceSet,
home: home,
}
return d, nil
}
@@ -58,7 +65,13 @@ func (d *Driver) Status() [][2]string {
}
func (d *Driver) Cleanup() error {
return d.DeviceSet.Shutdown()
err := d.DeviceSet.Shutdown()
if err2 := mount.Unmount(d.home); err == nil {
err = err2
}
return err
}
func (d *Driver) Create(id, parent string) error {
@@ -94,7 +107,7 @@ func (d *Driver) Get(id, mountLabel string) (string, error) {
mp := path.Join(d.home, "mnt", id)
// Create the target directories if they don't exist
if err := osMkdirAll(mp, 0755); err != nil && !osIsExist(err) {
if err := os.MkdirAll(mp, 0755); err != nil && !os.IsExist(err) {
return "", err
}
@@ -104,13 +117,13 @@ func (d *Driver) Get(id, mountLabel string) (string, error) {
}
rootFs := path.Join(mp, "rootfs")
if err := osMkdirAll(rootFs, 0755); err != nil && !osIsExist(err) {
if err := os.MkdirAll(rootFs, 0755); err != nil && !os.IsExist(err) {
d.DeviceSet.UnmountDevice(id)
return "", err
}
idFile := path.Join(mp, "id")
if _, err := osStat(idFile); err != nil && osIsNotExist(err) {
if _, err := os.Stat(idFile); err != nil && os.IsNotExist(err) {
// Create an "id" file with the container/image id in it to help reconscruct this in case
// of later problems
if err := ioutil.WriteFile(idFile, []byte(id), 0600); err != nil {
-880
View File
@@ -1,880 +0,0 @@
// +build linux,amd64
package devmapper
import (
"fmt"
"github.com/dotcloud/docker/daemon/graphdriver"
"io/ioutil"
"path"
"runtime"
"strings"
"syscall"
"testing"
)
func init() {
// Reduce the size the the base fs and loopback for the tests
DefaultDataLoopbackSize = 300 * 1024 * 1024
DefaultMetaDataLoopbackSize = 200 * 1024 * 1024
DefaultBaseFsSize = 300 * 1024 * 1024
}
// denyAllDevmapper mocks all calls to libdevmapper in the unit tests, and denies them by default
func denyAllDevmapper() {
// Hijack all calls to libdevmapper with default panics.
// Authorized calls are selectively hijacked in each tests.
DmTaskCreate = func(t int) *CDmTask {
panic("DmTaskCreate: this method should not be called here")
}
DmTaskRun = func(task *CDmTask) int {
panic("DmTaskRun: this method should not be called here")
}
DmTaskSetName = func(task *CDmTask, name string) int {
panic("DmTaskSetName: this method should not be called here")
}
DmTaskSetMessage = func(task *CDmTask, message string) int {
panic("DmTaskSetMessage: this method should not be called here")
}
DmTaskSetSector = func(task *CDmTask, sector uint64) int {
panic("DmTaskSetSector: this method should not be called here")
}
DmTaskSetCookie = func(task *CDmTask, cookie *uint, flags uint16) int {
panic("DmTaskSetCookie: this method should not be called here")
}
DmTaskSetAddNode = func(task *CDmTask, addNode AddNodeType) int {
panic("DmTaskSetAddNode: this method should not be called here")
}
DmTaskSetRo = func(task *CDmTask) int {
panic("DmTaskSetRo: this method should not be called here")
}
DmTaskAddTarget = func(task *CDmTask, start, size uint64, ttype, params string) int {
panic("DmTaskAddTarget: this method should not be called here")
}
DmTaskGetInfo = func(task *CDmTask, info *Info) int {
panic("DmTaskGetInfo: this method should not be called here")
}
DmGetNextTarget = func(task *CDmTask, next uintptr, start, length *uint64, target, params *string) uintptr {
panic("DmGetNextTarget: this method should not be called here")
}
DmUdevWait = func(cookie uint) int {
panic("DmUdevWait: this method should not be called here")
}
DmSetDevDir = func(dir string) int {
panic("DmSetDevDir: this method should not be called here")
}
DmGetLibraryVersion = func(version *string) int {
panic("DmGetLibraryVersion: this method should not be called here")
}
DmLogInitVerbose = func(level int) {
panic("DmLogInitVerbose: this method should not be called here")
}
DmTaskDestroy = func(task *CDmTask) {
panic("DmTaskDestroy: this method should not be called here")
}
LogWithErrnoInit = func() {
panic("LogWithErrnoInit: this method should not be called here")
}
}
func denyAllSyscall() {
sysMount = func(source, target, fstype string, flags uintptr, data string) (err error) {
panic("sysMount: this method should not be called here")
}
sysUnmount = func(target string, flags int) (err error) {
panic("sysUnmount: this method should not be called here")
}
sysCloseOnExec = func(fd int) {
panic("sysCloseOnExec: this method should not be called here")
}
sysSyscall = func(trap, a1, a2, a3 uintptr) (r1, r2 uintptr, err syscall.Errno) {
panic("sysSyscall: this method should not be called here")
}
// Not a syscall, but forbidding it here anyway
Mounted = func(mnt string) (bool, error) {
panic("devmapper.Mounted: this method should not be called here")
}
// osOpenFile = os.OpenFile
// osNewFile = os.NewFile
// osCreate = os.Create
// osStat = os.Stat
// osIsNotExist = os.IsNotExist
// osIsExist = os.IsExist
// osMkdirAll = os.MkdirAll
// osRemoveAll = os.RemoveAll
// osRename = os.Rename
// osReadlink = os.Readlink
// execRun = func(name string, args ...string) error {
// return exec.Command(name, args...).Run()
// }
}
func mkTestDirectory(t *testing.T) string {
dir, err := ioutil.TempDir("", "docker-test-devmapper-")
if err != nil {
t.Fatal(err)
}
return dir
}
func newDriver(t *testing.T) *Driver {
home := mkTestDirectory(t)
d, err := Init(home)
if err != nil {
t.Fatal(err)
}
return d.(*Driver)
}
func cleanup(d *Driver) {
d.Cleanup()
osRemoveAll(d.home)
}
type Set map[string]bool
func (r Set) Assert(t *testing.T, names ...string) {
for _, key := range names {
required := true
if strings.HasPrefix(key, "?") {
key = key[1:]
required = false
}
if _, exists := r[key]; !exists && required {
t.Fatalf("Key not set: %s", key)
}
delete(r, key)
}
if len(r) != 0 {
t.Fatalf("Unexpected keys: %v", r)
}
}
func TestInit(t *testing.T) {
var (
calls = make(Set)
taskMessages = make(Set)
taskTypes = make(Set)
home = mkTestDirectory(t)
)
defer osRemoveAll(home)
func() {
denyAllDevmapper()
DmSetDevDir = func(dir string) int {
calls["DmSetDevDir"] = true
expectedDir := "/dev"
if dir != expectedDir {
t.Fatalf("Wrong libdevmapper call\nExpected: DmSetDevDir(%v)\nReceived: DmSetDevDir(%v)\n", expectedDir, dir)
}
return 0
}
LogWithErrnoInit = func() {
calls["DmLogWithErrnoInit"] = true
}
var task1 CDmTask
DmTaskCreate = func(taskType int) *CDmTask {
calls["DmTaskCreate"] = true
taskTypes[fmt.Sprintf("%d", taskType)] = true
return &task1
}
DmTaskSetName = func(task *CDmTask, name string) int {
calls["DmTaskSetName"] = true
expectedTask := &task1
if task != expectedTask {
t.Fatalf("Wrong libdevmapper call\nExpected: DmTaskSetName(%v)\nReceived: DmTaskSetName(%v)\n", expectedTask, task)
}
// FIXME: use Set.AssertRegexp()
if !strings.HasPrefix(name, "docker-") && !strings.HasPrefix(name, "/dev/mapper/docker-") ||
!strings.HasSuffix(name, "-pool") && !strings.HasSuffix(name, "-base") {
t.Fatalf("Wrong libdevmapper call\nExpected: DmTaskSetName(%v)\nReceived: DmTaskSetName(%v)\n", "docker-...-pool", name)
}
return 1
}
DmTaskRun = func(task *CDmTask) int {
calls["DmTaskRun"] = true
expectedTask := &task1
if task != expectedTask {
t.Fatalf("Wrong libdevmapper call\nExpected: DmTaskRun(%v)\nReceived: DmTaskRun(%v)\n", expectedTask, task)
}
return 1
}
DmTaskGetInfo = func(task *CDmTask, info *Info) int {
calls["DmTaskGetInfo"] = true
expectedTask := &task1
if task != expectedTask {
t.Fatalf("Wrong libdevmapper call\nExpected: DmTaskGetInfo(%v)\nReceived: DmTaskGetInfo(%v)\n", expectedTask, task)
}
// This will crash if info is not dereferenceable
info.Exists = 0
return 1
}
DmTaskSetSector = func(task *CDmTask, sector uint64) int {
calls["DmTaskSetSector"] = true
expectedTask := &task1
if task != expectedTask {
t.Fatalf("Wrong libdevmapper call\nExpected: DmTaskSetSector(%v)\nReceived: DmTaskSetSector(%v)\n", expectedTask, task)
}
if expectedSector := uint64(0); sector != expectedSector {
t.Fatalf("Wrong libdevmapper call to DmTaskSetSector\nExpected: %v\nReceived: %v\n", expectedSector, sector)
}
return 1
}
DmTaskSetMessage = func(task *CDmTask, message string) int {
calls["DmTaskSetMessage"] = true
expectedTask := &task1
if task != expectedTask {
t.Fatalf("Wrong libdevmapper call\nExpected: DmTaskSetSector(%v)\nReceived: DmTaskSetSector(%v)\n", expectedTask, task)
}
taskMessages[message] = true
return 1
}
DmTaskDestroy = func(task *CDmTask) {
calls["DmTaskDestroy"] = true
expectedTask := &task1
if task != expectedTask {
t.Fatalf("Wrong libdevmapper call\nExpected: DmTaskDestroy(%v)\nReceived: DmTaskDestroy(%v)\n", expectedTask, task)
}
}
DmTaskAddTarget = func(task *CDmTask, start, size uint64, ttype, params string) int {
calls["DmTaskSetTarget"] = true
expectedTask := &task1
if task != expectedTask {
t.Fatalf("Wrong libdevmapper call\nExpected: DmTaskDestroy(%v)\nReceived: DmTaskDestroy(%v)\n", expectedTask, task)
}
if start != 0 {
t.Fatalf("Wrong start: %d != %d", start, 0)
}
if ttype != "thin" && ttype != "thin-pool" {
t.Fatalf("Wrong ttype: %s", ttype)
}
// Quick smoke test
if params == "" {
t.Fatalf("Params should not be empty")
}
return 1
}
fakeCookie := uint(4321)
DmTaskSetCookie = func(task *CDmTask, cookie *uint, flags uint16) int {
calls["DmTaskSetCookie"] = true
expectedTask := &task1
if task != expectedTask {
t.Fatalf("Wrong libdevmapper call\nExpected: DmTaskDestroy(%v)\nReceived: DmTaskDestroy(%v)\n", expectedTask, task)
}
if flags != 0 {
t.Fatalf("Cookie flags should be 0 (not %x)", flags)
}
*cookie = fakeCookie
return 1
}
DmUdevWait = func(cookie uint) int {
calls["DmUdevWait"] = true
if cookie != fakeCookie {
t.Fatalf("Wrong cookie: %d != %d", cookie, fakeCookie)
}
return 1
}
DmTaskSetAddNode = func(task *CDmTask, addNode AddNodeType) int {
if addNode != AddNodeOnCreate {
t.Fatalf("Wrong AddNoteType: %v (expected %v)", addNode, AddNodeOnCreate)
}
calls["DmTaskSetAddNode"] = true
return 1
}
execRun = func(name string, args ...string) error {
calls["execRun"] = true
if name != "mkfs.ext4" {
t.Fatalf("Expected %s to be executed, not %s", "mkfs.ext4", name)
}
return nil
}
driver, err := Init(home)
if err != nil {
t.Fatal(err)
}
defer func() {
if err := driver.Cleanup(); err != nil {
t.Fatal(err)
}
}()
}()
// Put all tests in a function to make sure the garbage collection will
// occur.
// Call GC to cleanup runtime.Finalizers
runtime.GC()
calls.Assert(t,
"DmSetDevDir",
"DmLogWithErrnoInit",
"DmTaskSetName",
"DmTaskRun",
"DmTaskGetInfo",
"DmTaskDestroy",
"execRun",
"DmTaskCreate",
"DmTaskSetTarget",
"DmTaskSetCookie",
"DmUdevWait",
"DmTaskSetSector",
"DmTaskSetMessage",
"DmTaskSetAddNode",
)
taskTypes.Assert(t, "0", "6", "17")
taskMessages.Assert(t, "create_thin 0", "set_transaction_id 0 1")
}
func fakeInit() func(home string) (graphdriver.Driver, error) {
oldInit := Init
Init = func(home string) (graphdriver.Driver, error) {
return &Driver{
home: home,
}, nil
}
return oldInit
}
func restoreInit(init func(home string) (graphdriver.Driver, error)) {
Init = init
}
func mockAllDevmapper(calls Set) {
DmSetDevDir = func(dir string) int {
calls["DmSetDevDir"] = true
return 0
}
LogWithErrnoInit = func() {
calls["DmLogWithErrnoInit"] = true
}
DmTaskCreate = func(taskType int) *CDmTask {
calls["DmTaskCreate"] = true
return &CDmTask{}
}
DmTaskSetName = func(task *CDmTask, name string) int {
calls["DmTaskSetName"] = true
return 1
}
DmTaskRun = func(task *CDmTask) int {
calls["DmTaskRun"] = true
return 1
}
DmTaskGetInfo = func(task *CDmTask, info *Info) int {
calls["DmTaskGetInfo"] = true
return 1
}
DmTaskSetSector = func(task *CDmTask, sector uint64) int {
calls["DmTaskSetSector"] = true
return 1
}
DmTaskSetMessage = func(task *CDmTask, message string) int {
calls["DmTaskSetMessage"] = true
return 1
}
DmTaskDestroy = func(task *CDmTask) {
calls["DmTaskDestroy"] = true
}
DmTaskAddTarget = func(task *CDmTask, start, size uint64, ttype, params string) int {
calls["DmTaskSetTarget"] = true
return 1
}
DmTaskSetCookie = func(task *CDmTask, cookie *uint, flags uint16) int {
calls["DmTaskSetCookie"] = true
return 1
}
DmUdevWait = func(cookie uint) int {
calls["DmUdevWait"] = true
return 1
}
DmTaskSetAddNode = func(task *CDmTask, addNode AddNodeType) int {
calls["DmTaskSetAddNode"] = true
return 1
}
execRun = func(name string, args ...string) error {
calls["execRun"] = true
return nil
}
}
func TestDriverName(t *testing.T) {
denyAllDevmapper()
defer denyAllDevmapper()
oldInit := fakeInit()
defer restoreInit(oldInit)
d := newDriver(t)
if d.String() != "devicemapper" {
t.Fatalf("Expected driver name to be devicemapper got %s", d.String())
}
}
func TestDriverCreate(t *testing.T) {
denyAllDevmapper()
denyAllSyscall()
defer denyAllSyscall()
defer denyAllDevmapper()
calls := make(Set)
mockAllDevmapper(calls)
sysMount = func(source, target, fstype string, flags uintptr, data string) (err error) {
calls["sysMount"] = true
// FIXME: compare the exact source and target strings (inodes + devname)
if expectedSource := "/dev/mapper/docker-"; !strings.HasPrefix(source, expectedSource) {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedSource, source)
}
if expectedTarget := "/tmp/docker-test-devmapper-"; !strings.HasPrefix(target, expectedTarget) {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedTarget, target)
}
if expectedFstype := "ext4"; fstype != expectedFstype {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedFstype, fstype)
}
if expectedFlags := uintptr(3236757504); flags != expectedFlags {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedFlags, flags)
}
return nil
}
sysUnmount = func(target string, flag int) error {
//calls["sysUnmount"] = true
return nil
}
Mounted = func(mnt string) (bool, error) {
calls["Mounted"] = true
if !strings.HasPrefix(mnt, "/tmp/docker-test-devmapper-") || !strings.HasSuffix(mnt, "/mnt/1") {
t.Fatalf("Wrong mounted call\nExpected: Mounted(%v)\nReceived: Mounted(%v)\n", "/tmp/docker-test-devmapper-.../mnt/1", mnt)
}
return false, nil
}
sysSyscall = func(trap, a1, a2, a3 uintptr) (r1, r2 uintptr, err syscall.Errno) {
calls["sysSyscall"] = true
if trap != sysSysIoctl {
t.Fatalf("Unexpected syscall. Expecting SYS_IOCTL, received: %d", trap)
}
switch a2 {
case LoopSetFd:
calls["ioctl.loopsetfd"] = true
case LoopCtlGetFree:
calls["ioctl.loopctlgetfree"] = true
case LoopGetStatus64:
calls["ioctl.loopgetstatus"] = true
case LoopSetStatus64:
calls["ioctl.loopsetstatus"] = true
case LoopClrFd:
calls["ioctl.loopclrfd"] = true
case LoopSetCapacity:
calls["ioctl.loopsetcapacity"] = true
case BlkGetSize64:
calls["ioctl.blkgetsize"] = true
default:
t.Fatalf("Unexpected IOCTL. Received %d", a2)
}
return 0, 0, 0
}
func() {
d := newDriver(t)
calls.Assert(t,
"DmSetDevDir",
"DmLogWithErrnoInit",
"DmTaskSetName",
"DmTaskRun",
"DmTaskGetInfo",
"execRun",
"DmTaskCreate",
"DmTaskSetTarget",
"DmTaskSetCookie",
"DmUdevWait",
"DmTaskSetSector",
"DmTaskSetMessage",
"DmTaskSetAddNode",
"sysSyscall",
"ioctl.blkgetsize",
"ioctl.loopsetfd",
"ioctl.loopsetstatus",
"?ioctl.loopctlgetfree",
)
if err := d.Create("1", ""); err != nil {
t.Fatal(err)
}
calls.Assert(t,
"DmTaskCreate",
"DmTaskGetInfo",
"DmTaskRun",
"DmTaskSetSector",
"DmTaskSetName",
"DmTaskSetMessage",
)
}()
runtime.GC()
calls.Assert(t,
"DmTaskDestroy",
)
}
func TestDriverRemove(t *testing.T) {
denyAllDevmapper()
denyAllSyscall()
defer denyAllSyscall()
defer denyAllDevmapper()
calls := make(Set)
mockAllDevmapper(calls)
sysMount = func(source, target, fstype string, flags uintptr, data string) (err error) {
calls["sysMount"] = true
// FIXME: compare the exact source and target strings (inodes + devname)
if expectedSource := "/dev/mapper/docker-"; !strings.HasPrefix(source, expectedSource) {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedSource, source)
}
if expectedTarget := "/tmp/docker-test-devmapper-"; !strings.HasPrefix(target, expectedTarget) {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedTarget, target)
}
if expectedFstype := "ext4"; fstype != expectedFstype {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedFstype, fstype)
}
if expectedFlags := uintptr(3236757504); flags != expectedFlags {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedFlags, flags)
}
return nil
}
sysUnmount = func(target string, flags int) (err error) {
// FIXME: compare the exact source and target strings (inodes + devname)
if expectedTarget := "/tmp/docker-test-devmapper-"; !strings.HasPrefix(target, expectedTarget) {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedTarget, target)
}
if expectedFlags := 0; flags != expectedFlags {
t.Fatalf("Wrong syscall call\nExpected: Mount(%v)\nReceived: Mount(%v)\n", expectedFlags, flags)
}
return nil
}
Mounted = func(mnt string) (bool, error) {
calls["Mounted"] = true
return false, nil
}
sysSyscall = func(trap, a1, a2, a3 uintptr) (r1, r2 uintptr, err syscall.Errno) {
calls["sysSyscall"] = true
if trap != sysSysIoctl {
t.Fatalf("Unexpected syscall. Expecting SYS_IOCTL, received: %d", trap)
}
switch a2 {
case LoopSetFd:
calls["ioctl.loopsetfd"] = true
case LoopCtlGetFree:
calls["ioctl.loopctlgetfree"] = true
case LoopGetStatus64:
calls["ioctl.loopgetstatus"] = true
case LoopSetStatus64:
calls["ioctl.loopsetstatus"] = true
case LoopClrFd:
calls["ioctl.loopclrfd"] = true
case LoopSetCapacity:
calls["ioctl.loopsetcapacity"] = true
case BlkGetSize64:
calls["ioctl.blkgetsize"] = true
default:
t.Fatalf("Unexpected IOCTL. Received %d", a2)
}
return 0, 0, 0
}
func() {
d := newDriver(t)
calls.Assert(t,
"DmSetDevDir",
"DmLogWithErrnoInit",
"DmTaskSetName",
"DmTaskRun",
"DmTaskGetInfo",
"execRun",
"DmTaskCreate",
"DmTaskSetTarget",
"DmTaskSetCookie",
"DmUdevWait",
"DmTaskSetSector",
"DmTaskSetMessage",
"DmTaskSetAddNode",
"sysSyscall",
"ioctl.blkgetsize",
"ioctl.loopsetfd",
"ioctl.loopsetstatus",
"?ioctl.loopctlgetfree",
)
if err := d.Create("1", ""); err != nil {
t.Fatal(err)
}
calls.Assert(t,
"DmTaskCreate",
"DmTaskGetInfo",
"DmTaskRun",
"DmTaskSetSector",
"DmTaskSetName",
"DmTaskSetMessage",
)
Mounted = func(mnt string) (bool, error) {
calls["Mounted"] = true
return true, nil
}
if err := d.Remove("1"); err != nil {
t.Fatal(err)
}
calls.Assert(t,
"DmTaskRun",
"DmTaskSetSector",
"DmTaskSetName",
"DmTaskSetMessage",
"DmTaskCreate",
"DmTaskGetInfo",
"DmTaskSetCookie",
"DmTaskSetTarget",
"DmTaskSetAddNode",
"DmUdevWait",
)
}()
runtime.GC()
calls.Assert(t,
"DmTaskDestroy",
)
}
func TestCleanup(t *testing.T) {
t.Skip("FIXME: not a unit test")
t.Skip("Unimplemented")
d := newDriver(t)
defer osRemoveAll(d.home)
mountPoints := make([]string, 2)
if err := d.Create("1", ""); err != nil {
t.Fatal(err)
}
// Mount the id
p, err := d.Get("1", "")
if err != nil {
t.Fatal(err)
}
mountPoints[0] = p
if err := d.Create("2", "1"); err != nil {
t.Fatal(err)
}
p, err = d.Get("2", "")
if err != nil {
t.Fatal(err)
}
mountPoints[1] = p
// Ensure that all the mount points are currently mounted
for _, p := range mountPoints {
if mounted, err := Mounted(p); err != nil {
t.Fatal(err)
} else if !mounted {
t.Fatalf("Expected %s to be mounted", p)
}
}
// Ensure that devices are active
for _, p := range []string{"1", "2"} {
if !d.HasActivatedDevice(p) {
t.Fatalf("Expected %s to have an active device", p)
}
}
if err := d.Cleanup(); err != nil {
t.Fatal(err)
}
// Ensure that all the mount points are no longer mounted
for _, p := range mountPoints {
if mounted, err := Mounted(p); err != nil {
t.Fatal(err)
} else if mounted {
t.Fatalf("Expected %s to not be mounted", p)
}
}
// Ensure that devices are no longer activated
for _, p := range []string{"1", "2"} {
if d.HasActivatedDevice(p) {
t.Fatalf("Expected %s not be an active device", p)
}
}
}
func TestNotMounted(t *testing.T) {
t.Skip("FIXME: not a unit test")
t.Skip("Not implemented")
d := newDriver(t)
defer cleanup(d)
if err := d.Create("1", ""); err != nil {
t.Fatal(err)
}
mounted, err := Mounted(path.Join(d.home, "mnt", "1"))
if err != nil {
t.Fatal(err)
}
if mounted {
t.Fatal("Id 1 should not be mounted")
}
}
func TestMounted(t *testing.T) {
t.Skip("FIXME: not a unit test")
d := newDriver(t)
defer cleanup(d)
if err := d.Create("1", ""); err != nil {
t.Fatal(err)
}
if _, err := d.Get("1", ""); err != nil {
t.Fatal(err)
}
mounted, err := Mounted(path.Join(d.home, "mnt", "1"))
if err != nil {
t.Fatal(err)
}
if !mounted {
t.Fatal("Id 1 should be mounted")
}
}
func TestInitCleanedDriver(t *testing.T) {
t.Skip("FIXME: not a unit test")
d := newDriver(t)
if err := d.Create("1", ""); err != nil {
t.Fatal(err)
}
if _, err := d.Get("1", ""); err != nil {
t.Fatal(err)
}
if err := d.Cleanup(); err != nil {
t.Fatal(err)
}
driver, err := Init(d.home)
if err != nil {
t.Fatal(err)
}
d = driver.(*Driver)
defer cleanup(d)
if _, err := d.Get("1", ""); err != nil {
t.Fatal(err)
}
}
func TestMountMountedDriver(t *testing.T) {
t.Skip("FIXME: not a unit test")
d := newDriver(t)
defer cleanup(d)
if err := d.Create("1", ""); err != nil {
t.Fatal(err)
}
// Perform get on same id to ensure that it will
// not be mounted twice
if _, err := d.Get("1", ""); err != nil {
t.Fatal(err)
}
if _, err := d.Get("1", ""); err != nil {
t.Fatal(err)
}
}
func TestGetReturnsValidDevice(t *testing.T) {
t.Skip("FIXME: not a unit test")
d := newDriver(t)
defer cleanup(d)
if err := d.Create("1", ""); err != nil {
t.Fatal(err)
}
if !d.HasDevice("1") {
t.Fatalf("Expected id 1 to be in device set")
}
if _, err := d.Get("1", ""); err != nil {
t.Fatal(err)
}
if !d.HasActivatedDevice("1") {
t.Fatalf("Expected id 1 to be activated")
}
if !d.HasInitializedDevice("1") {
t.Fatalf("Expected id 1 to be initialized")
}
}
func TestDriverGetSize(t *testing.T) {
t.Skip("FIXME: not a unit test")
t.Skipf("Size is currently not implemented")
d := newDriver(t)
defer cleanup(d)
if err := d.Create("1", ""); err != nil {
t.Fatal(err)
}
mountPoint, err := d.Get("1", "")
if err != nil {
t.Fatal(err)
}
size := int64(1024)
f, err := osCreate(path.Join(mountPoint, "test_file"))
if err != nil {
t.Fatal(err)
}
if err := f.Truncate(size); err != nil {
t.Fatal(err)
}
f.Close()
// diffSize, err := d.DiffSize("1")
// if err != nil {
// t.Fatal(err)
// }
// if diffSize != size {
// t.Fatalf("Expected size %d got %d", size, diffSize)
// }
}
func assertMap(t *testing.T, m map[string]bool, keys ...string) {
for _, key := range keys {
if _, exists := m[key]; !exists {
t.Fatalf("Key not set: %s", key)
}
delete(m, key)
}
if len(m) != 0 {
t.Fatalf("Unexpected keys: %v", m)
}
}
+9 -8
View File
@@ -3,11 +3,12 @@
package devmapper
import (
"syscall"
"unsafe"
)
func ioctlLoopCtlGetFree(fd uintptr) (int, error) {
index, _, err := sysSyscall(sysSysIoctl, fd, LoopCtlGetFree, 0)
index, _, err := syscall.Syscall(syscall.SYS_IOCTL, fd, LoopCtlGetFree, 0)
if err != 0 {
return 0, err
}
@@ -15,21 +16,21 @@ func ioctlLoopCtlGetFree(fd uintptr) (int, error) {
}
func ioctlLoopSetFd(loopFd, sparseFd uintptr) error {
if _, _, err := sysSyscall(sysSysIoctl, loopFd, LoopSetFd, sparseFd); err != 0 {
if _, _, err := syscall.Syscall(syscall.SYS_IOCTL, loopFd, LoopSetFd, sparseFd); err != 0 {
return err
}
return nil
}
func ioctlLoopSetStatus64(loopFd uintptr, loopInfo *LoopInfo64) error {
if _, _, err := sysSyscall(sysSysIoctl, loopFd, LoopSetStatus64, uintptr(unsafe.Pointer(loopInfo))); err != 0 {
if _, _, err := syscall.Syscall(syscall.SYS_IOCTL, loopFd, LoopSetStatus64, uintptr(unsafe.Pointer(loopInfo))); err != 0 {
return err
}
return nil
}
func ioctlLoopClrFd(loopFd uintptr) error {
if _, _, err := sysSyscall(sysSysIoctl, loopFd, LoopClrFd, 0); err != 0 {
if _, _, err := syscall.Syscall(syscall.SYS_IOCTL, loopFd, LoopClrFd, 0); err != 0 {
return err
}
return nil
@@ -38,14 +39,14 @@ func ioctlLoopClrFd(loopFd uintptr) error {
func ioctlLoopGetStatus64(loopFd uintptr) (*LoopInfo64, error) {
loopInfo := &LoopInfo64{}
if _, _, err := sysSyscall(sysSysIoctl, loopFd, LoopGetStatus64, uintptr(unsafe.Pointer(loopInfo))); err != 0 {
if _, _, err := syscall.Syscall(syscall.SYS_IOCTL, loopFd, LoopGetStatus64, uintptr(unsafe.Pointer(loopInfo))); err != 0 {
return nil, err
}
return loopInfo, nil
}
func ioctlLoopSetCapacity(loopFd uintptr, value int) error {
if _, _, err := sysSyscall(sysSysIoctl, loopFd, LoopSetCapacity, uintptr(value)); err != 0 {
if _, _, err := syscall.Syscall(syscall.SYS_IOCTL, loopFd, LoopSetCapacity, uintptr(value)); err != 0 {
return err
}
return nil
@@ -53,7 +54,7 @@ func ioctlLoopSetCapacity(loopFd uintptr, value int) error {
func ioctlBlkGetSize64(fd uintptr) (int64, error) {
var size int64
if _, _, err := sysSyscall(sysSysIoctl, fd, BlkGetSize64, uintptr(unsafe.Pointer(&size))); err != 0 {
if _, _, err := syscall.Syscall(syscall.SYS_IOCTL, fd, BlkGetSize64, uintptr(unsafe.Pointer(&size))); err != 0 {
return 0, err
}
return size, nil
@@ -64,7 +65,7 @@ func ioctlBlkDiscard(fd uintptr, offset, length uint64) error {
r[0] = offset
r[1] = length
if _, _, err := sysSyscall(sysSysIoctl, fd, BlkDiscard, uintptr(unsafe.Pointer(&r[0]))); err != 0 {
if _, _, err := syscall.Syscall(syscall.SYS_IOCTL, fd, BlkDiscard, uintptr(unsafe.Pointer(&r[0]))); err != 0 {
return err
}
return nil
+65 -6
View File
@@ -3,25 +3,84 @@
package devmapper
import (
"bytes"
"fmt"
"os"
"path/filepath"
"syscall"
)
// FIXME: this is copy-pasted from the aufs driver.
// It should be moved into the core.
var Mounted = func(mountpoint string) (bool, error) {
mntpoint, err := osStat(mountpoint)
func Mounted(mountpoint string) (bool, error) {
mntpoint, err := os.Stat(mountpoint)
if err != nil {
if osIsNotExist(err) {
if os.IsNotExist(err) {
return false, nil
}
return false, err
}
parent, err := osStat(filepath.Join(mountpoint, ".."))
parent, err := os.Stat(filepath.Join(mountpoint, ".."))
if err != nil {
return false, err
}
mntpointSt := toSysStatT(mntpoint.Sys())
parentSt := toSysStatT(parent.Sys())
mntpointSt := mntpoint.Sys().(*syscall.Stat_t)
parentSt := parent.Sys().(*syscall.Stat_t)
return mntpointSt.Dev != parentSt.Dev, nil
}
type probeData struct {
fsName string
magic string
offset uint64
}
func ProbeFsType(device string) (string, error) {
probes := []probeData{
{"btrfs", "_BHRfS_M", 0x10040},
{"ext4", "\123\357", 0x438},
{"xfs", "XFSB", 0},
}
maxLen := uint64(0)
for _, p := range probes {
l := p.offset + uint64(len(p.magic))
if l > maxLen {
maxLen = l
}
}
file, err := os.Open(device)
if err != nil {
return "", err
}
buffer := make([]byte, maxLen)
l, err := file.Read(buffer)
if err != nil {
return "", err
}
file.Close()
if uint64(l) != maxLen {
return "", fmt.Errorf("unable to detect filesystem type of %s, short read", device)
}
for _, p := range probes {
if bytes.Equal([]byte(p.magic), buffer[p.offset:p.offset+uint64(len(p.magic))]) {
return p.fsName, nil
}
}
return "", fmt.Errorf("Unknown filesystem type on %s", device)
}
func joinMountOptions(a, b string) string {
if a == "" {
return b
}
if b == "" {
return a
}
return a + "," + b
}
-57
View File
@@ -1,57 +0,0 @@
// +build linux,amd64
package devmapper
import (
"os"
"os/exec"
"syscall"
)
type (
sysStatT syscall.Stat_t
sysErrno syscall.Errno
osFile struct{ *os.File }
)
var (
sysMount = syscall.Mount
sysUnmount = syscall.Unmount
sysCloseOnExec = syscall.CloseOnExec
sysSyscall = syscall.Syscall
osOpenFile = func(name string, flag int, perm os.FileMode) (*osFile, error) {
f, err := os.OpenFile(name, flag, perm)
return &osFile{File: f}, err
}
osOpen = func(name string) (*osFile, error) { f, err := os.Open(name); return &osFile{File: f}, err }
osNewFile = os.NewFile
osCreate = os.Create
osStat = os.Stat
osIsNotExist = os.IsNotExist
osIsExist = os.IsExist
osMkdirAll = os.MkdirAll
osRemoveAll = os.RemoveAll
osRename = os.Rename
osReadlink = os.Readlink
execRun = func(name string, args ...string) error { return exec.Command(name, args...).Run() }
)
const (
sysMsMgcVal = syscall.MS_MGC_VAL
sysMsRdOnly = syscall.MS_RDONLY
sysEInval = syscall.EINVAL
sysSysIoctl = syscall.SYS_IOCTL
sysEBusy = syscall.EBUSY
osORdOnly = os.O_RDONLY
osORdWr = os.O_RDWR
osOCreate = os.O_CREATE
osModeDevice = os.ModeDevice
)
func toSysStatT(i interface{}) *sysStatT {
return (*sysStatT)(i.(*syscall.Stat_t))
}
+48 -14
View File
@@ -1,14 +1,23 @@
package graphdriver
import (
"errors"
"fmt"
"github.com/dotcloud/docker/archive"
"github.com/dotcloud/docker/utils"
"os"
"path"
"github.com/dotcloud/docker/archive"
"github.com/dotcloud/docker/pkg/mount"
)
type InitFunc func(root string) (Driver, error)
type FsMagic uint64
const (
FsMagicBtrfs = FsMagic(0x9123683E)
FsMagicAufs = FsMagic(0x61756673)
)
type InitFunc func(root string, options []string) (Driver, error)
type Driver interface {
String() string
@@ -43,6 +52,10 @@ var (
"devicemapper",
"vfs",
}
ErrNotSupported = errors.New("driver not supported")
ErrPrerequisites = errors.New("prerequisites for driver not satisfied (wrong filesystem?)")
ErrIncompatibleFS = fmt.Errorf("backing file system is unsupported for this graph driver")
)
func init() {
@@ -58,35 +71,56 @@ func Register(name string, initFunc InitFunc) error {
return nil
}
func GetDriver(name, home string) (Driver, error) {
func GetDriver(name, home string, options []string) (Driver, error) {
if initFunc, exists := drivers[name]; exists {
return initFunc(path.Join(home, name))
return initFunc(path.Join(home, name), options)
}
return nil, fmt.Errorf("No such driver: %s", name)
return nil, ErrNotSupported
}
func New(root string) (driver Driver, err error) {
func New(root string, options []string) (driver Driver, err error) {
for _, name := range []string{os.Getenv("DOCKER_DRIVER"), DefaultDriver} {
if name != "" {
return GetDriver(name, root)
return GetDriver(name, root, options)
}
}
// Check for priority drivers first
for _, name := range priority {
if driver, err = GetDriver(name, root); err != nil {
utils.Debugf("Error loading driver %s: %s", name, err)
continue
driver, err = GetDriver(name, root, options)
if err != nil {
if err == ErrNotSupported || err == ErrPrerequisites || err == ErrIncompatibleFS {
continue
}
return nil, err
}
return driver, nil
}
// Check all registered drivers if no priority driver is found
for _, initFunc := range drivers {
if driver, err = initFunc(root); err != nil {
continue
if driver, err = initFunc(root, options); err != nil {
if err == ErrNotSupported || err == ErrPrerequisites || err == ErrIncompatibleFS {
continue
}
return nil, err
}
return driver, nil
}
return nil, err
return nil, fmt.Errorf("No supported storage backend found")
}
func MakePrivate(mountPoint string) error {
mounted, err := mount.Mounted(mountPoint)
if err != nil {
return err
}
if !mounted {
if err := mount.Mount(mountPoint, mountPoint, "none", "bind,rw"); err != nil {
return err
}
}
return mount.ForceMount("", mountPoint, "none", "private")
}
+228
View File
@@ -0,0 +1,228 @@
package graphtest
import (
"github.com/dotcloud/docker/daemon/graphdriver"
"io/ioutil"
"os"
"path"
"syscall"
"testing"
)
var (
drv *Driver
)
type Driver struct {
graphdriver.Driver
root string
refCount int
}
func newDriver(t *testing.T, name string) *Driver {
root, err := ioutil.TempDir("/var/tmp", "docker-graphtest-")
if err != nil {
t.Fatal(err)
}
if err := os.MkdirAll(root, 0755); err != nil {
t.Fatal(err)
}
d, err := graphdriver.GetDriver(name, root, nil)
if err != nil {
if err == graphdriver.ErrNotSupported || err == graphdriver.ErrPrerequisites {
t.Skip("Driver %s not supported", name)
}
t.Fatal(err)
}
return &Driver{d, root, 1}
}
func cleanup(t *testing.T, d *Driver) {
if err := drv.Cleanup(); err != nil {
t.Fatal(err)
}
os.RemoveAll(d.root)
}
func GetDriver(t *testing.T, name string) graphdriver.Driver {
if drv == nil {
drv = newDriver(t, name)
} else {
drv.refCount++
}
return drv
}
func PutDriver(t *testing.T) {
if drv == nil {
t.Skip("No driver to put!")
}
drv.refCount--
if drv.refCount == 0 {
cleanup(t, drv)
drv = nil
}
}
func verifyFile(t *testing.T, path string, mode os.FileMode, uid, gid uint32) {
fi, err := os.Stat(path)
if err != nil {
t.Fatal(err)
}
if fi.Mode()&os.ModeType != mode&os.ModeType {
t.Fatalf("Expected %s type 0x%x, got 0x%x", path, mode&os.ModeType, fi.Mode()&os.ModeType)
}
if fi.Mode()&os.ModePerm != mode&os.ModePerm {
t.Fatalf("Expected %s mode %o, got %o", path, mode&os.ModePerm, fi.Mode()&os.ModePerm)
}
if fi.Mode()&os.ModeSticky != mode&os.ModeSticky {
t.Fatalf("Expected %s sticky 0x%x, got 0x%x", path, mode&os.ModeSticky, fi.Mode()&os.ModeSticky)
}
if fi.Mode()&os.ModeSetuid != mode&os.ModeSetuid {
t.Fatalf("Expected %s setuid 0x%x, got 0x%x", path, mode&os.ModeSetuid, fi.Mode()&os.ModeSetuid)
}
if fi.Mode()&os.ModeSetgid != mode&os.ModeSetgid {
t.Fatalf("Expected %s setgid 0x%x, got 0x%x", path, mode&os.ModeSetgid, fi.Mode()&os.ModeSetgid)
}
if stat, ok := fi.Sys().(*syscall.Stat_t); ok {
if stat.Uid != uid {
t.Fatal("%s no owned by uid %d", path, uid)
}
if stat.Gid != gid {
t.Fatal("%s not owned by gid %d", path, gid)
}
}
}
// Creates an new image and verifies it is empty and the right metadata
func DriverTestCreateEmpty(t *testing.T, drivername string) {
driver := GetDriver(t, drivername)
defer PutDriver(t)
if err := driver.Create("empty", ""); err != nil {
t.Fatal(err)
}
if !driver.Exists("empty") {
t.Fatal("Newly created image doesn't exist")
}
dir, err := driver.Get("empty", "")
if err != nil {
t.Fatal(err)
}
verifyFile(t, dir, 0755|os.ModeDir, 0, 0)
// Verify that the directory is empty
fis, err := ioutil.ReadDir(dir)
if err != nil {
t.Fatal(err)
}
if len(fis) != 0 {
t.Fatal("New directory not empty")
}
driver.Put("empty")
if err := driver.Remove("empty"); err != nil {
t.Fatal(err)
}
}
func createBase(t *testing.T, driver graphdriver.Driver, name string) {
// We need to be able to set any perms
oldmask := syscall.Umask(0)
defer syscall.Umask(oldmask)
if err := driver.Create(name, ""); err != nil {
t.Fatal(err)
}
dir, err := driver.Get(name, "")
if err != nil {
t.Fatal(err)
}
defer driver.Put(name)
subdir := path.Join(dir, "a subdir")
if err := os.Mkdir(subdir, 0705|os.ModeSticky); err != nil {
t.Fatal(err)
}
if err := os.Chown(subdir, 1, 2); err != nil {
t.Fatal(err)
}
file := path.Join(dir, "a file")
if err := ioutil.WriteFile(file, []byte("Some data"), 0222|os.ModeSetuid); err != nil {
t.Fatal(err)
}
}
func verifyBase(t *testing.T, driver graphdriver.Driver, name string) {
dir, err := driver.Get(name, "")
if err != nil {
t.Fatal(err)
}
defer driver.Put(name)
subdir := path.Join(dir, "a subdir")
verifyFile(t, subdir, 0705|os.ModeDir|os.ModeSticky, 1, 2)
file := path.Join(dir, "a file")
verifyFile(t, file, 0222|os.ModeSetuid, 0, 0)
fis, err := ioutil.ReadDir(dir)
if err != nil {
t.Fatal(err)
}
if len(fis) != 2 {
t.Fatal("Unexpected files in base image")
}
}
func DriverTestCreateBase(t *testing.T, drivername string) {
driver := GetDriver(t, drivername)
defer PutDriver(t)
createBase(t, driver, "Base")
verifyBase(t, driver, "Base")
if err := driver.Remove("Base"); err != nil {
t.Fatal(err)
}
}
func DriverTestCreateSnap(t *testing.T, drivername string) {
driver := GetDriver(t, drivername)
defer PutDriver(t)
createBase(t, driver, "Base")
if err := driver.Create("Snap", "Base"); err != nil {
t.Fatal(err)
}
verifyBase(t, driver, "Snap")
if err := driver.Remove("Snap"); err != nil {
t.Fatal(err)
}
if err := driver.Remove("Base"); err != nil {
t.Fatal(err)
}
}
+2 -2
View File
@@ -12,7 +12,7 @@ func init() {
graphdriver.Register("vfs", Init)
}
func Init(home string) (graphdriver.Driver, error) {
func Init(home string, options []string) (graphdriver.Driver, error) {
d := &Driver{
home: home,
}
@@ -47,7 +47,7 @@ func (d *Driver) Create(id, parent string) error {
if err := os.MkdirAll(path.Dir(dir), 0700); err != nil {
return err
}
if err := os.Mkdir(dir, 0700); err != nil {
if err := os.Mkdir(dir, 0755); err != nil {
return err
}
if parent == "" {
+28
View File
@@ -0,0 +1,28 @@
package vfs
import (
"github.com/dotcloud/docker/daemon/graphdriver/graphtest"
"testing"
)
// This avoids creating a new driver for each test if all tests are run
// Make sure to put new tests between TestVfsSetup and TestVfsTeardown
func TestVfsSetup(t *testing.T) {
graphtest.GetDriver(t, "vfs")
}
func TestVfsCreateEmpty(t *testing.T) {
graphtest.DriverTestCreateEmpty(t, "vfs")
}
func TestVfsCreateBase(t *testing.T) {
graphtest.DriverTestCreateBase(t, "vfs")
}
func TestVfsCreateSnap(t *testing.T) {
graphtest.DriverTestCreateSnap(t, "vfs")
}
func TestVfsTeardown(t *testing.T) {
graphtest.PutDriver(t)
}
+3
View File
@@ -26,5 +26,8 @@ func (history *History) Swap(i, j int) {
func (history *History) Add(container *Container) {
*history = append(*history, container)
}
func (history *History) Sort() {
sort.Sort(history)
}
+54
View File
@@ -0,0 +1,54 @@
package daemon
import (
"encoding/json"
"github.com/dotcloud/docker/engine"
"github.com/dotcloud/docker/runconfig"
)
func (daemon *Daemon) ContainerInspect(job *engine.Job) engine.Status {
if len(job.Args) != 1 {
return job.Errorf("usage: %s NAME", job.Name)
}
name := job.Args[0]
if container := daemon.Get(name); container != nil {
if job.GetenvBool("dirty") {
b, err := json.Marshal(&struct {
*Container
HostConfig *runconfig.HostConfig
}{container, container.HostConfig()})
if err != nil {
return job.Error(err)
}
job.Stdout.Write(b)
return engine.StatusOK
}
out := &engine.Env{}
out.Set("Id", container.ID)
out.SetAuto("Created", container.Created)
out.Set("Path", container.Path)
out.SetList("Args", container.Args)
out.SetJson("Config", container.Config)
out.SetJson("State", container.State)
out.Set("Image", container.Image)
out.SetJson("NetworkSettings", container.NetworkSettings)
out.Set("ResolvConfPath", container.ResolvConfPath)
out.Set("HostnamePath", container.HostnamePath)
out.Set("HostsPath", container.HostsPath)
out.Set("Name", container.Name)
out.Set("Driver", container.Driver)
out.Set("ExecDriver", container.ExecDriver)
out.Set("MountLabel", container.MountLabel)
out.Set("ProcessLabel", container.ProcessLabel)
out.SetJson("Volumes", container.Volumes)
out.SetJson("VolumesRW", container.VolumesRW)
out.SetJson("HostConfig", container.hostConfig)
if _, err := out.WriteTo(job.Stdout); err != nil {
return job.Error(err)
}
return engine.StatusOK
}
return job.Errorf("No such container: %s", name)
}
+1 -1
View File
@@ -23,7 +23,7 @@ func (settings *NetworkSettings) PortMappingAPI() *engine.Table {
p, _ := nat.ParsePort(port.Port())
if len(bindings) == 0 {
out := &engine.Env{}
out.SetInt("PublicPort", p)
out.SetInt("PrivatePort", p)
out.Set("Type", port.Proto())
outs.Add(out)
continue
+55 -20
View File
@@ -6,6 +6,7 @@ import (
"log"
"net"
"strings"
"sync"
"github.com/dotcloud/docker/daemon/networkdriver"
"github.com/dotcloud/docker/daemon/networkdriver/ipallocator"
@@ -28,6 +29,24 @@ type networkInterface struct {
PortMappings []net.Addr // there are mappings to the host interfaces
}
type ifaces struct {
c map[string]*networkInterface
sync.Mutex
}
func (i *ifaces) Set(key string, n *networkInterface) {
i.Lock()
i.c[key] = n
i.Unlock()
}
func (i *ifaces) Get(key string) *networkInterface {
i.Lock()
res := i.c[key]
i.Unlock()
return res
}
var (
addrs = []string{
// Here we don't follow the convention of using the 1st IP of the range for the gateway.
@@ -53,7 +72,7 @@ var (
bridgeNetwork *net.IPNet
defaultBindingIP = net.ParseIP("0.0.0.0")
currentInterfaces = make(map[string]*networkInterface)
currentInterfaces = ifaces{c: make(map[string]*networkInterface)}
)
func InitDriver(job *engine.Job) engine.Status {
@@ -321,9 +340,9 @@ func Allocate(job *engine.Job) engine.Status {
size, _ := bridgeNetwork.Mask.Size()
out.SetInt("IPPrefixLen", size)
currentInterfaces[id] = &networkInterface{
currentInterfaces.Set(id, &networkInterface{
IP: *ip,
}
})
out.WriteTo(job.Stdout)
@@ -334,7 +353,7 @@ func Allocate(job *engine.Job) engine.Status {
func Release(job *engine.Job) engine.Status {
var (
id = job.Args[0]
containerInterface = currentInterfaces[id]
containerInterface = currentInterfaces.Get(id)
ip net.IP
port int
proto string
@@ -380,39 +399,55 @@ func AllocatePort(job *engine.Job) engine.Status {
ip = defaultBindingIP
id = job.Args[0]
hostIP = job.Getenv("HostIP")
hostPort = job.GetenvInt("HostPort")
origHostPort = job.GetenvInt("HostPort")
containerPort = job.GetenvInt("ContainerPort")
proto = job.Getenv("Proto")
network = currentInterfaces[id]
network = currentInterfaces.Get(id)
)
if hostIP != "" {
ip = net.ParseIP(hostIP)
}
// host ip, proto, and host port
hostPort, err = portallocator.RequestPort(ip, proto, hostPort)
if err != nil {
return job.Error(err)
}
var (
hostPort int
container net.Addr
host net.Addr
)
if proto == "tcp" {
host = &net.TCPAddr{IP: ip, Port: hostPort}
container = &net.TCPAddr{IP: network.IP, Port: containerPort}
} else {
host = &net.UDPAddr{IP: ip, Port: hostPort}
container = &net.UDPAddr{IP: network.IP, Port: containerPort}
/*
Try up to 10 times to get a port that's not already allocated.
In the event of failure to bind, return the error that portmapper.Map
yields.
*/
for i := 0; i < 10; i++ {
// host ip, proto, and host port
hostPort, err = portallocator.RequestPort(ip, proto, origHostPort)
if err != nil {
return job.Error(err)
}
if proto == "tcp" {
host = &net.TCPAddr{IP: ip, Port: hostPort}
container = &net.TCPAddr{IP: network.IP, Port: containerPort}
} else {
host = &net.UDPAddr{IP: ip, Port: hostPort}
container = &net.UDPAddr{IP: network.IP, Port: containerPort}
}
if err = portmapper.Map(container, ip, hostPort); err == nil {
break
}
job.Logf("Failed to bind %s:%d for container address %s:%d. Trying another port.", ip.String(), hostPort, network.IP.String(), containerPort)
}
if err := portmapper.Map(container, ip, hostPort); err != nil {
portallocator.ReleasePort(ip, proto, hostPort)
if err != nil {
return job.Error(err)
}
network.PortMappings = append(network.PortMappings, host)
out := engine.Env{}
+51 -83
View File
@@ -4,12 +4,21 @@ import (
"encoding/binary"
"errors"
"github.com/dotcloud/docker/daemon/networkdriver"
"github.com/dotcloud/docker/pkg/collections"
"net"
"sync"
)
type networkSet map[string]*collections.OrderedIntSet
// allocatedMap is thread-unsafe set of allocated IP
type allocatedMap struct {
p map[int32]struct{}
last int32
}
func newAllocatedMap() *allocatedMap {
return &allocatedMap{p: make(map[int32]struct{})}
}
type networkSet map[string]*allocatedMap
var (
ErrNoAvailableIPs = errors.New("no available ip addresses on network")
@@ -19,92 +28,74 @@ var (
var (
lock = sync.Mutex{}
allocatedIPs = networkSet{}
availableIPS = networkSet{}
)
// RequestIP requests an available ip from the given network. It
// will return the next available ip if the ip provided is nil. If the
// ip provided is not nil it will validate that the provided ip is available
// for use or return an error
func RequestIP(address *net.IPNet, ip *net.IP) (*net.IP, error) {
func RequestIP(network *net.IPNet, ip *net.IP) (*net.IP, error) {
lock.Lock()
defer lock.Unlock()
checkAddress(address)
key := network.String()
allocated, ok := allocatedIPs[key]
if !ok {
allocated = newAllocatedMap()
allocatedIPs[key] = allocated
}
if ip == nil {
next, err := getNextIp(address)
if err != nil {
return nil, err
}
return next, nil
return allocated.getNextIP(network)
}
if err := registerIP(address, ip); err != nil {
return nil, err
}
return ip, nil
return allocated.checkIP(network, ip)
}
// ReleaseIP adds the provided ip back into the pool of
// available ips to be returned for use.
func ReleaseIP(address *net.IPNet, ip *net.IP) error {
func ReleaseIP(network *net.IPNet, ip *net.IP) error {
lock.Lock()
defer lock.Unlock()
checkAddress(address)
var (
existing = allocatedIPs[address.String()]
available = availableIPS[address.String()]
pos = getPosition(address, ip)
)
existing.Remove(int(pos))
available.Push(int(pos))
if allocated, exists := allocatedIPs[network.String()]; exists {
pos := getPosition(network, ip)
delete(allocated.p, pos)
}
return nil
}
// convert the ip into the position in the subnet. Only
// position are saved in the set
func getPosition(address *net.IPNet, ip *net.IP) int32 {
var (
first, _ = networkdriver.NetworkRange(address)
base = ipToInt(&first)
i = ipToInt(ip)
)
return i - base
func getPosition(network *net.IPNet, ip *net.IP) int32 {
first, _ := networkdriver.NetworkRange(network)
return ipToInt(ip) - ipToInt(&first)
}
func (allocated *allocatedMap) checkIP(network *net.IPNet, ip *net.IP) (*net.IP, error) {
pos := getPosition(network, ip)
if _, ok := allocated.p[pos]; ok {
return nil, ErrIPAlreadyAllocated
}
allocated.p[pos] = struct{}{}
allocated.last = pos
return ip, nil
}
// return an available ip if one is currently available. If not,
// return the next available ip for the nextwork
func getNextIp(address *net.IPNet) (*net.IP, error) {
func (allocated *allocatedMap) getNextIP(network *net.IPNet) (*net.IP, error) {
var (
ownIP = ipToInt(&address.IP)
available = availableIPS[address.String()]
allocated = allocatedIPs[address.String()]
first, _ = networkdriver.NetworkRange(address)
base = ipToInt(&first)
size = int(networkdriver.NetworkSize(address.Mask))
max = int32(size - 2) // size -1 for the broadcast address, -1 for the gateway address
pos = int32(available.Pop())
ownIP = ipToInt(&network.IP)
first, _ = networkdriver.NetworkRange(network)
base = ipToInt(&first)
size = int(networkdriver.NetworkSize(network.Mask))
max = int32(size - 2) // size -1 for the broadcast network, -1 for the gateway network
pos = allocated.last
)
// We pop and push the position not the ip
if pos != 0 {
ip := intToIP(int32(base + pos))
allocated.Push(int(pos))
return ip, nil
}
var (
firstNetIP = address.IP.To4().Mask(address.Mask)
firstNetIP = network.IP.To4().Mask(network.Mask)
firstAsInt = ipToInt(&firstNetIP) + 1
)
pos = int32(allocated.PullBack())
for i := int32(0); i < max; i++ {
pos = pos%max + 1
next := int32(base + pos)
@@ -112,31 +103,16 @@ func getNextIp(address *net.IPNet) (*net.IP, error) {
if next == ownIP || next == firstAsInt {
continue
}
if !allocated.Exists(int(pos)) {
ip := intToIP(next)
allocated.Push(int(pos))
return ip, nil
if _, ok := allocated.p[pos]; ok {
continue
}
allocated.p[pos] = struct{}{}
allocated.last = pos
return intToIP(next), nil
}
return nil, ErrNoAvailableIPs
}
func registerIP(address *net.IPNet, ip *net.IP) error {
var (
existing = allocatedIPs[address.String()]
available = availableIPS[address.String()]
pos = getPosition(address, ip)
)
if existing.Exists(int(pos)) {
return ErrIPAlreadyAllocated
}
available.Remove(int(pos))
return nil
}
// Converts a 4 bytes IP into a 32 bit integer
func ipToInt(ip *net.IP) int32 {
return int32(binary.BigEndian.Uint32(ip.To4()))
@@ -149,11 +125,3 @@ func intToIP(n int32) *net.IP {
ip := net.IP(b)
return &ip
}
func checkAddress(address *net.IPNet) {
key := address.String()
if _, exists := allocatedIPs[key]; !exists {
allocatedIPs[key] = collections.NewOrderedIntSet()
availableIPS[key] = collections.NewOrderedIntSet()
}
}
@@ -8,7 +8,6 @@ import (
func reset() {
allocatedIPs = networkSet{}
availableIPS = networkSet{}
}
func TestRequestNewIps(t *testing.T) {
@@ -18,8 +17,10 @@ func TestRequestNewIps(t *testing.T) {
Mask: []byte{255, 255, 255, 0},
}
var ip *net.IP
var err error
for i := 2; i < 10; i++ {
ip, err := RequestIP(network, nil)
ip, err = RequestIP(network, nil)
if err != nil {
t.Fatal(err)
}
@@ -28,6 +29,17 @@ func TestRequestNewIps(t *testing.T) {
t.Fatalf("Expected ip %s got %s", expected, ip.String())
}
}
value := intToIP(ipToInt(ip) + 1).String()
if err := ReleaseIP(network, ip); err != nil {
t.Fatal(err)
}
ip, err = RequestIP(network, nil)
if err != nil {
t.Fatal(err)
}
if ip.String() != value {
t.Fatalf("Expected to receive the next ip %s got %s", value, ip.String())
}
}
func TestReleaseIp(t *testing.T) {
@@ -64,6 +76,17 @@ func TestGetReleasedIp(t *testing.T) {
t.Fatal(err)
}
for i := 0; i < 252; i++ {
_, err = RequestIP(network, nil)
if err != nil {
t.Fatal(err)
}
err = ReleaseIP(network, ip)
if err != nil {
t.Fatal(err)
}
}
ip, err = RequestIP(network, nil)
if err != nil {
t.Fatal(err)
@@ -185,24 +208,6 @@ func TestIPAllocator(t *testing.T) {
newIPs[i] = ip
}
// Before loop begin
// 2(u) - 3(u) - 4(f) - 5(f) - 6(f)
// ↑
// After i = 0
// 2(u) - 3(u) - 4(f) - 5(u) - 6(f)
// ↑
// After i = 1
// 2(u) - 3(u) - 4(f) - 5(u) - 6(u)
// ↑
// After i = 2
// 2(u) - 3(u) - 4(u) - 5(u) - 6(u)
// ↑
// Reordered these because the new set will always return the
// lowest ips first and not in the order that they were released
assertIPEquals(t, &expectedIPs[2], newIPs[0])
assertIPEquals(t, &expectedIPs[3], newIPs[1])
assertIPEquals(t, &expectedIPs[4], newIPs[2])
@@ -234,8 +239,105 @@ func TestAllocateFirstIP(t *testing.T) {
}
}
func TestAllocateAllIps(t *testing.T) {
defer reset()
network := &net.IPNet{
IP: []byte{192, 168, 0, 1},
Mask: []byte{255, 255, 255, 0},
}
var (
current, first *net.IP
err error
isFirst = true
)
for err == nil {
current, err = RequestIP(network, nil)
if isFirst {
first = current
isFirst = false
}
}
if err != ErrNoAvailableIPs {
t.Fatal(err)
}
if _, err := RequestIP(network, nil); err != ErrNoAvailableIPs {
t.Fatal(err)
}
if err := ReleaseIP(network, first); err != nil {
t.Fatal(err)
}
again, err := RequestIP(network, nil)
if err != nil {
t.Fatal(err)
}
assertIPEquals(t, first, again)
}
func TestAllocateDifferentSubnets(t *testing.T) {
defer reset()
network1 := &net.IPNet{
IP: []byte{192, 168, 0, 1},
Mask: []byte{255, 255, 255, 0},
}
network2 := &net.IPNet{
IP: []byte{127, 0, 0, 1},
Mask: []byte{255, 255, 255, 0},
}
expectedIPs := []net.IP{
0: net.IPv4(192, 168, 0, 2),
1: net.IPv4(192, 168, 0, 3),
2: net.IPv4(127, 0, 0, 2),
3: net.IPv4(127, 0, 0, 3),
}
ip11, err := RequestIP(network1, nil)
if err != nil {
t.Fatal(err)
}
ip12, err := RequestIP(network1, nil)
if err != nil {
t.Fatal(err)
}
ip21, err := RequestIP(network2, nil)
if err != nil {
t.Fatal(err)
}
ip22, err := RequestIP(network2, nil)
if err != nil {
t.Fatal(err)
}
assertIPEquals(t, &expectedIPs[0], ip11)
assertIPEquals(t, &expectedIPs[1], ip12)
assertIPEquals(t, &expectedIPs[2], ip21)
assertIPEquals(t, &expectedIPs[3], ip22)
}
func assertIPEquals(t *testing.T, ip1, ip2 *net.IP) {
if !ip1.Equal(*ip2) {
t.Fatalf("Expected IP %s, got %s", ip1, ip2)
}
}
func BenchmarkRequestIP(b *testing.B) {
network := &net.IPNet{
IP: []byte{192, 168, 0, 1},
Mask: []byte{255, 255, 255, 0},
}
b.ResetTimer()
for i := 0; i < b.N; i++ {
for j := 0; j < 253; j++ {
_, err := RequestIP(network, nil)
if err != nil {
b.Fatal(err)
}
}
reset()
}
}
@@ -2,21 +2,21 @@ package portallocator
import (
"errors"
"github.com/dotcloud/docker/pkg/collections"
"net"
"sync"
)
type (
portMap map[int]bool
protocolMap map[string]portMap
ipMapping map[string]protocolMap
)
const (
BeginPortRange = 49153
EndPortRange = 65535
)
type (
portMappings map[string]*collections.OrderedIntSet
ipMapping map[string]portMappings
)
var (
ErrAllPortsAllocated = errors.New("all ports are allocated")
ErrPortAlreadyAllocated = errors.New("port has already been allocated")
@@ -24,165 +24,106 @@ var (
)
var (
currentDynamicPort = map[string]int{
"tcp": BeginPortRange - 1,
"udp": BeginPortRange - 1,
}
defaultIP = net.ParseIP("0.0.0.0")
defaultAllocatedPorts = portMappings{}
otherAllocatedPorts = ipMapping{}
lock = sync.Mutex{}
mutex sync.Mutex
defaultIP = net.ParseIP("0.0.0.0")
globalMap = ipMapping{}
)
func init() {
defaultAllocatedPorts["tcp"] = collections.NewOrderedIntSet()
defaultAllocatedPorts["udp"] = collections.NewOrderedIntSet()
}
// RequestPort returns an available port if the port is 0
// If the provided port is not 0 then it will be checked if
// it is available for allocation
func RequestPort(ip net.IP, proto string, port int) (int, error) {
lock.Lock()
defer lock.Unlock()
mutex.Lock()
defer mutex.Unlock()
if err := validateProtocol(proto); err != nil {
if err := validateProto(proto); err != nil {
return 0, err
}
// If the user requested a specific port to be allocated
ip = getDefault(ip)
mapping := getOrCreate(ip)
if port > 0 {
if err := registerSetPort(ip, proto, port); err != nil {
if !mapping[proto][port] {
mapping[proto][port] = true
return port, nil
} else {
return 0, ErrPortAlreadyAllocated
}
} else {
port, err := findPort(ip, proto)
if err != nil {
return 0, err
}
return port, nil
}
return registerDynamicPort(ip, proto)
}
// ReleasePort will return the provided port back into the
// pool for reuse
func ReleasePort(ip net.IP, proto string, port int) error {
lock.Lock()
defer lock.Unlock()
mutex.Lock()
defer mutex.Unlock()
if err := validateProtocol(proto); err != nil {
return err
}
ip = getDefault(ip)
allocated := defaultAllocatedPorts[proto]
allocated.Remove(port)
mapping := getOrCreate(ip)
delete(mapping[proto], port)
if !equalsDefault(ip) {
registerIP(ip)
// Remove the port for the specific ip address
allocated = otherAllocatedPorts[ip.String()][proto]
allocated.Remove(port)
}
return nil
}
func ReleaseAll() error {
lock.Lock()
defer lock.Unlock()
mutex.Lock()
defer mutex.Unlock()
currentDynamicPort["tcp"] = BeginPortRange - 1
currentDynamicPort["udp"] = BeginPortRange - 1
defaultAllocatedPorts = portMappings{}
defaultAllocatedPorts["tcp"] = collections.NewOrderedIntSet()
defaultAllocatedPorts["udp"] = collections.NewOrderedIntSet()
otherAllocatedPorts = ipMapping{}
globalMap = ipMapping{}
return nil
}
func registerDynamicPort(ip net.IP, proto string) (int, error) {
func getOrCreate(ip net.IP) protocolMap {
ipstr := ip.String()
if !equalsDefault(ip) {
registerIP(ip)
ipAllocated := otherAllocatedPorts[ip.String()][proto]
port, err := findNextPort(proto, ipAllocated)
if err != nil {
return 0, err
if _, ok := globalMap[ipstr]; !ok {
globalMap[ipstr] = protocolMap{
"tcp": portMap{},
"udp": portMap{},
}
ipAllocated.Push(port)
return port, nil
} else {
allocated := defaultAllocatedPorts[proto]
port, err := findNextPort(proto, allocated)
if err != nil {
return 0, err
}
allocated.Push(port)
return port, nil
}
}
func registerSetPort(ip net.IP, proto string, port int) error {
allocated := defaultAllocatedPorts[proto]
if allocated.Exists(port) {
return ErrPortAlreadyAllocated
}
if !equalsDefault(ip) {
registerIP(ip)
ipAllocated := otherAllocatedPorts[ip.String()][proto]
if ipAllocated.Exists(port) {
return ErrPortAlreadyAllocated
}
ipAllocated.Push(port)
} else {
allocated.Push(port)
}
return nil
return globalMap[ipstr]
}
func equalsDefault(ip net.IP) bool {
return ip == nil || ip.Equal(defaultIP)
}
func findPort(ip net.IP, proto string) (int, error) {
port := BeginPortRange
func findNextPort(proto string, allocated *collections.OrderedIntSet) (int, error) {
port := nextPort(proto)
startSearchPort := port
for allocated.Exists(port) {
port = nextPort(proto)
if startSearchPort == port {
mapping := getOrCreate(ip)
for mapping[proto][port] {
port++
if port > EndPortRange {
return 0, ErrAllPortsAllocated
}
}
mapping[proto][port] = true
return port, nil
}
func nextPort(proto string) int {
c := currentDynamicPort[proto] + 1
if c > EndPortRange {
c = BeginPortRange
func getDefault(ip net.IP) net.IP {
if ip == nil {
return defaultIP
}
currentDynamicPort[proto] = c
return c
return ip
}
func registerIP(ip net.IP) {
if _, exists := otherAllocatedPorts[ip.String()]; !exists {
otherAllocatedPorts[ip.String()] = portMappings{
"tcp": collections.NewOrderedIntSet(),
"udp": collections.NewOrderedIntSet(),
}
}
}
func validateProtocol(proto string) error {
if _, exists := defaultAllocatedPorts[proto]; !exists {
func validateProto(proto string) error {
if proto != "tcp" && proto != "udp" {
return ErrUnknownProtocol
}
return nil
}
+28 -3
View File
@@ -2,14 +2,16 @@ package daemon
import (
"fmt"
"github.com/dotcloud/docker/utils"
"sync"
"time"
"github.com/dotcloud/docker/pkg/units"
)
type State struct {
sync.RWMutex
Running bool
Paused bool
Pid int
ExitCode int
StartedAt time.Time
@@ -22,12 +24,15 @@ func (s *State) String() string {
defer s.RUnlock()
if s.Running {
return fmt.Sprintf("Up %s", utils.HumanDuration(time.Now().UTC().Sub(s.StartedAt)))
if s.Paused {
return fmt.Sprintf("Up %s (Paused)", units.HumanDuration(time.Now().UTC().Sub(s.StartedAt)))
}
return fmt.Sprintf("Up %s", units.HumanDuration(time.Now().UTC().Sub(s.StartedAt)))
}
if s.FinishedAt.IsZero() {
return ""
}
return fmt.Sprintf("Exited (%d) %s ago", s.ExitCode, utils.HumanDuration(time.Now().UTC().Sub(s.FinishedAt)))
return fmt.Sprintf("Exited (%d) %s ago", s.ExitCode, units.HumanDuration(time.Now().UTC().Sub(s.FinishedAt)))
}
func (s *State) IsRunning() bool {
@@ -49,6 +54,7 @@ func (s *State) SetRunning(pid int) {
defer s.Unlock()
s.Running = true
s.Paused = false
s.ExitCode = 0
s.Pid = pid
s.StartedAt = time.Now().UTC()
@@ -63,3 +69,22 @@ func (s *State) SetStopped(exitCode int) {
s.FinishedAt = time.Now().UTC()
s.ExitCode = exitCode
}
func (s *State) SetPaused() {
s.Lock()
defer s.Unlock()
s.Paused = true
}
func (s *State) SetUnpaused() {
s.Lock()
defer s.Unlock()
s.Paused = false
}
func (s *State) IsPaused() bool {
s.RLock()
defer s.RUnlock()
return s.Paused
}
+3 -16
View File
@@ -2,10 +2,10 @@ package daemon
import (
"fmt"
"github.com/dotcloud/docker/nat"
"github.com/dotcloud/docker/pkg/namesgenerator"
"github.com/dotcloud/docker/runconfig"
"strings"
"github.com/dotcloud/docker/nat"
"github.com/dotcloud/docker/runconfig"
)
func migratePortMappings(config *runconfig.Config, hostConfig *runconfig.HostConfig) error {
@@ -49,16 +49,3 @@ func mergeLxcConfIntoOptions(hostConfig *runconfig.HostConfig, driverConfig map[
driverConfig["lxc"] = lxc
}
}
type checker struct {
daemon *Daemon
}
func (c *checker) Exists(name string) bool {
return c.daemon.containerGraph.Exists("/" + name)
}
// Generate a random and unique name
func generateRandomName(daemon *Daemon) (string, error) {
return namesgenerator.GenerateRandomName(&checker{daemon})
}
+160 -122
View File
@@ -10,7 +10,7 @@ import (
"github.com/dotcloud/docker/archive"
"github.com/dotcloud/docker/daemon/execdriver"
"github.com/dotcloud/docker/utils"
"github.com/dotcloud/docker/pkg/symlink"
)
type BindMap struct {
@@ -40,8 +40,11 @@ func setupMountsForContainer(container *Container) error {
{container.ResolvConfPath, "/etc/resolv.conf", false, true},
}
if container.HostnamePath != "" && container.HostsPath != "" {
if container.HostnamePath != "" {
mounts = append(mounts, execdriver.Mount{container.HostnamePath, "/etc/hostname", false, true})
}
if container.HostsPath != "" {
mounts = append(mounts, execdriver.Mount{container.HostsPath, "/etc/hosts", false, true})
}
@@ -94,13 +97,16 @@ func applyVolumesFrom(container *Container) error {
if _, exists := container.Volumes[volPath]; exists {
continue
}
stat, err := os.Stat(filepath.Join(c.basefs, volPath))
stat, err := os.Stat(c.getResourcePath(volPath))
if err != nil {
return err
}
if err := createIfNotExists(filepath.Join(container.basefs, volPath), stat.IsDir()); err != nil {
if err := createIfNotExists(container.getResourcePath(volPath), stat.IsDir()); err != nil {
return err
}
container.Volumes[volPath] = id
if isRW, exists := c.VolumesRW[volPath]; exists {
container.VolumesRW[volPath] = isRW && mountRW
@@ -162,137 +168,169 @@ func createVolumes(container *Container) error {
return err
}
volumesDriver := container.daemon.volumes.Driver()
// Create the requested volumes if they don't exist
for volPath := range container.Config.Volumes {
volPath = filepath.Clean(volPath)
volIsDir := true
// Skip existing volumes
if _, exists := container.Volumes[volPath]; exists {
continue
}
var srcPath string
var isBindMount bool
srcRW := false
// If an external bind is defined for this volume, use that as a source
if bindMap, exists := binds[volPath]; exists {
isBindMount = true
srcPath = bindMap.SrcPath
if !filepath.IsAbs(srcPath) {
return fmt.Errorf("%s must be an absolute path", srcPath)
}
if strings.ToLower(bindMap.Mode) == "rw" {
srcRW = true
}
if stat, err := os.Stat(bindMap.SrcPath); err != nil {
return err
} else {
volIsDir = stat.IsDir()
}
// Otherwise create an directory in $ROOT/volumes/ and use that
} else {
// Do not pass a container as the parameter for the volume creation.
// The graph driver using the container's information ( Image ) to
// create the parent.
c, err := container.daemon.volumes.Create(nil, "", "", "", "", nil, nil)
if err != nil {
return err
}
srcPath, err = volumesDriver.Get(c.ID, "")
if err != nil {
return fmt.Errorf("Driver %s failed to get volume rootfs %s: %s", volumesDriver, c.ID, err)
}
srcRW = true // RW by default
}
if p, err := filepath.EvalSymlinks(srcPath); err != nil {
return err
} else {
srcPath = p
}
// Create the mountpoint
rootVolPath, err := utils.FollowSymlinkInScope(filepath.Join(container.basefs, volPath), container.basefs)
if err != nil {
if err := initializeVolume(container, volPath, binds); err != nil {
return err
}
}
newVolPath, err := filepath.Rel(container.basefs, rootVolPath)
if err != nil {
for volPath := range binds {
if err := initializeVolume(container, volPath, binds); err != nil {
return err
}
newVolPath = "/" + newVolPath
if volPath != newVolPath {
delete(container.Volumes, volPath)
delete(container.VolumesRW, volPath)
}
container.Volumes[newVolPath] = srcPath
container.VolumesRW[newVolPath] = srcRW
if err := createIfNotExists(rootVolPath, volIsDir); err != nil {
return err
}
// Do not copy or change permissions if we are mounting from the host
if srcRW && !isBindMount {
volList, err := ioutil.ReadDir(rootVolPath)
if err != nil {
return err
}
if len(volList) > 0 {
srcList, err := ioutil.ReadDir(srcPath)
if err != nil {
return err
}
if len(srcList) == 0 {
// If the source volume is empty copy files from the root into the volume
if err := archive.CopyWithTar(rootVolPath, srcPath); err != nil {
return err
}
}
}
var stat syscall.Stat_t
if err := syscall.Stat(rootVolPath, &stat); err != nil {
return err
}
var srcStat syscall.Stat_t
if err := syscall.Stat(srcPath, &srcStat); err != nil {
return err
}
// Change the source volume's ownership if it differs from the root
// files that were just copied
if stat.Uid != srcStat.Uid || stat.Gid != srcStat.Gid {
if err := os.Chown(srcPath, int(stat.Uid), int(stat.Gid)); err != nil {
return err
}
}
}
}
return nil
}
func createIfNotExists(path string, isDir bool) error {
if _, err := os.Stat(path); err != nil {
if os.IsNotExist(err) {
if isDir {
if err := os.MkdirAll(path, 0755); err != nil {
return err
}
} else {
if err := os.MkdirAll(filepath.Dir(path), 0755); err != nil {
return err
}
f, err := os.OpenFile(path, os.O_CREATE, 0755)
if err != nil {
return err
}
defer f.Close()
func createIfNotExists(destination string, isDir bool) error {
if _, err := os.Stat(destination); err != nil && os.IsNotExist(err) {
if isDir {
if err := os.MkdirAll(destination, 0755); err != nil {
return err
}
} else {
if err := os.MkdirAll(filepath.Dir(destination), 0755); err != nil {
return err
}
f, err := os.OpenFile(destination, os.O_CREATE, 0755)
if err != nil {
return err
}
f.Close()
}
}
return nil
}
func initializeVolume(container *Container, volPath string, binds map[string]BindMap) error {
volumesDriver := container.daemon.volumes.Driver()
volPath = filepath.Clean(volPath)
// Skip existing volumes
if _, exists := container.Volumes[volPath]; exists {
return nil
}
var (
destination string
isBindMount bool
volIsDir = true
srcRW = false
)
// If an external bind is defined for this volume, use that as a source
if bindMap, exists := binds[volPath]; exists {
isBindMount = true
destination = bindMap.SrcPath
if !filepath.IsAbs(destination) {
return fmt.Errorf("%s must be an absolute path", destination)
}
if strings.ToLower(bindMap.Mode) == "rw" {
srcRW = true
}
if stat, err := os.Stat(bindMap.SrcPath); err != nil {
return err
} else {
volIsDir = stat.IsDir()
}
} else {
// Do not pass a container as the parameter for the volume creation.
// The graph driver using the container's information ( Image ) to
// create the parent.
c, err := container.daemon.volumes.Create(nil, "", "", "", "", nil, nil)
if err != nil {
return err
}
destination, err = volumesDriver.Get(c.ID, "")
if err != nil {
return fmt.Errorf("Driver %s failed to get volume rootfs %s: %s", volumesDriver, c.ID, err)
}
srcRW = true
}
if p, err := filepath.EvalSymlinks(destination); err != nil {
return err
} else {
destination = p
}
// Create the mountpoint
source, err := symlink.FollowSymlinkInScope(filepath.Join(container.basefs, volPath), container.basefs)
if err != nil {
return err
}
newVolPath, err := filepath.Rel(container.basefs, source)
if err != nil {
return err
}
newVolPath = "/" + newVolPath
if volPath != newVolPath {
delete(container.Volumes, volPath)
delete(container.VolumesRW, volPath)
}
container.Volumes[newVolPath] = destination
container.VolumesRW[newVolPath] = srcRW
if err := createIfNotExists(source, volIsDir); err != nil {
return err
}
// Do not copy or change permissions if we are mounting from the host
if srcRW && !isBindMount {
if err := copyExistingContents(source, destination); err != nil {
return err
}
}
return nil
}
func copyExistingContents(source, destination string) error {
volList, err := ioutil.ReadDir(source)
if err != nil {
return err
}
if len(volList) > 0 {
srcList, err := ioutil.ReadDir(destination)
if err != nil {
return err
}
if len(srcList) == 0 {
// If the source volume is empty copy files from the root into the volume
if err := archive.CopyWithTar(source, destination); err != nil {
return err
}
}
}
return copyOwnership(source, destination)
}
// copyOwnership copies the permissions and uid:gid of the source file
// into the destination file
func copyOwnership(source, destination string) error {
var stat syscall.Stat_t
if err := syscall.Stat(source, &stat); err != nil {
return err
}
if err := os.Chown(destination, int(stat.Uid), int(stat.Gid)); err != nil {
return err
}
return os.Chmod(destination, os.FileMode(stat.Mode))
}
+3
View File
@@ -0,0 +1,3 @@
This directory contains code pertaining to the configuration of the docker deamon
These are the configuration settings that you pass to the docker daemon when you launch it with say: `docker -d -e lxc`
+5
View File
@@ -25,6 +25,7 @@ type Config struct {
BridgeIP string
InterContainerCommunication bool
GraphDriver string
GraphOptions []string
ExecDriver string
Mtu int
DisableNetwork bool
@@ -49,6 +50,10 @@ func ConfigFromJob(job *engine.Job) *Config {
ExecDriver: job.Getenv("ExecDriver"),
EnableSelinuxSupport: job.GetenvBool("EnableSelinuxSupport"),
}
if graphOpts := job.GetenvList("GraphOptions"); graphOpts != nil {
config.GraphOptions = graphOpts
}
if dns := job.GetenvList("Dns"); dns != nil {
config.Dns = dns
}
+7
View File
@@ -41,6 +41,7 @@ func main() {
var (
flVersion = flag.Bool([]string{"v", "-version"}, false, "Print version information and quit")
flDaemon = flag.Bool([]string{"d", "-daemon"}, false, "Enable daemon mode")
flGraphOpts opts.ListOpts
flDebug = flag.Bool([]string{"D", "-debug"}, false, "Enable debug mode")
flAutoRestart = flag.Bool([]string{"r", "-restart"}, true, "Restart previously running containers")
bridgeName = flag.String([]string{"b", "-bridge"}, "", "Attach containers to a pre-existing network bridge\nuse 'none' to disable container networking")
@@ -69,6 +70,7 @@ func main() {
flag.Var(&flDns, []string{"#dns", "-dns"}, "Force docker to use specific DNS servers")
flag.Var(&flDnsSearch, []string{"-dns-search"}, "Force Docker to use specific DNS search domains")
flag.Var(&flHosts, []string{"H", "-host"}, "The socket(s) to bind to in daemon mode\nspecified using one or more tcp://host:port, unix:///path/to/socket, fd://* or fd://socketfd.")
flag.Var(&flGraphOpts, []string{"-storage-opt"}, "Set storage driver options")
flag.Parse()
@@ -98,6 +100,9 @@ func main() {
}
if *flDaemon {
if runtime.GOOS != "linux" {
log.Fatalf("The Docker daemon is only supported on linux")
}
if os.Geteuid() != 0 {
log.Fatalf("The Docker daemon needs to be run as root")
}
@@ -153,6 +158,7 @@ func main() {
job.Setenv("DefaultIp", *flDefaultIp)
job.SetenvBool("InterContainerCommunication", *flInterContainerComm)
job.Setenv("GraphDriver", *flGraphDriver)
job.SetenvList("GraphOptions", flGraphOpts.GetAll())
job.Setenv("ExecDriver", *flExecDriver)
job.SetenvInt("Mtu", *flMtu)
job.SetenvBool("EnableSelinuxSupport", *flSelinuxEnabled)
@@ -185,6 +191,7 @@ func main() {
job.Setenv("TlsCa", *flCa)
job.Setenv("TlsCert", *flCert)
job.Setenv("TlsKey", *flKey)
job.SetenvBool("BufferRequests", true)
if err := job.Run(); err != nil {
log.Fatal(err)
}

Some files were not shown because too many files have changed in this diff Show More