move the description of user ns mapping and default files to proper file #171

gao-feng · 2015-09-10T10:10:52Z

They should stay in runtime not config.

Signed-off-by: Gao feng omarapazanadi@gmail.com

wking · 2015-09-10T16:31:28Z

runtime-config.md

@@ -36,6 +36,15 @@ Only [mounts from the portable config](config.md#mount-points) will be mounted.
    }
 }
 ```
+Note: The following filesystems MUST be made available in each Linux application's filesystem


This is Linux-specific stuff, so I think it should go in the Linux-specific docs too.

@wking I thinks describe this after the mounts Exmaple for linux will exactly tell people to make sure these filesystems related mounts should be configured in runtime.json

On Thu, Sep 10, 2015 at 08:46:15PM -0700, Gao feng wrote:

+Note: The following filesystems MUST be made available in each Linux application's filesystem

@wking I thinks describe this after the mounts Exmaple for linux…

That's a good point. I guess my preferred approach would be to move
the Linux-specific mount example into the Linux-specific
documentation, and just link to the per-platform examples from this
platform-agnostic documentation. Otherwise this is a reasonable chunk
of noise for folks working on other platforms.

Good idea, will update, thanks!

wking · 2015-09-10T16:35:47Z

On Thu, Sep 10, 2015 at 03:10:53AM -0700, Gao feng wrote:

They should stay in runtime not config.

I'd rather land #126 (which needs a rebase now, but I'm waiting until
there's enough interest in merging it before doing that ;) or at least
some of its earlier commits. With an easier mapping between JSON
files and their associated .md and .go, it's more obvious when docs
end up in the wrong place.

gao-feng · 2015-09-11T14:47:30Z

@wking Updated, please check, thanks!

wking · 2015-09-11T15:03:34Z

runtime-config-linux.md

+*hostID* is the starting uid/gid on the host to be mapped to *containerID* which is the starting uid/gid in the container and *size* refers to the number of ids to be mapped.
+There is a limit of 5 mappings which is the Linux kernel hard limit.
+
+```json


I know you're just copy-pasting here, but since you reorder, it might be worth adding the *Example* psuedo-header that seems to be our convention for setting off examples.

@wking You mean we should add Example before all of json examples in this file?

On Sun, Sep 13, 2015 at 07:34:11PM -0700, Gao feng wrote:

@wking You mean we should add Example before all of json examples
in this file?

Yes, although we can leave examples you aren't touching for follow-up
PRs (although it looks like you got a lot of them in 107a454, and
I'm fine with that too).

Thanks, append another commit :)

463042f looks good to me.

mrunalp · 2015-10-05T20:01:15Z

config-linux.md

-*hostID* is the starting uid/gid on the host to be mapped to *containerID* which is the starting uid/gid in the container and *size* refers to the number of ids to be mapped.
-There is a limit of 5 mappings which is the Linux kernel hard limit.
-
-## Default Devices and File Systems


This section could remain here as it is talking about the mounts and not the runtime details.

On Mon, Oct 05, 2015 at 01:01:17PM -0700, Mrunal Patel wrote:

-## Default Devices and File Systems

This section could remain here as it is talking about the mounts and
not the runtime details.

Currently devices are a purely runtime.json issue (although see 1,
where I suggest moving the mknod bits back into config.json). So I'm
ok keeping the default mounts here, but the default devices seem like
they should be moved into the runtime.json docs.

mrunalp · 2015-10-05T20:02:02Z

@gao-feng Could you update and rebase the PR?

gao-feng · 2015-10-09T04:10:49Z

@mrunalp updated, please check.

vishh · 2015-12-16T18:17:27Z

LGTM

vbatts · 2015-12-16T20:30:44Z

rebase and LGTM

I filed a PR to keep this separate [1], but it was closed after [2] landed. See also [3], where I point out that putting the mknod stuff in the “control groups” section is awkward. [1]: opencontainers/runtime-spec#99 Add linux.resources.devices [2]: opencontainers/runtime-spec#94 Replace Linux.Device with more specific config [3]: opencontainers/runtime-spec#171 (comment) move the description of user ns mapping and default files to proper file

Signed-off-by: Gao feng <omarapazanadi@gmail.com>

They should stay in runtime not config. Signed-off-by: Gao feng <omarapazanadi@gmail.com>

gao-feng · 2016-01-05T06:23:15Z

@wking @mrunalp @vishh @vbatts rebased, please check :)

hqhq · 2016-01-05T06:53:45Z

LGTM

move the description of user ns mapping and default files to proper file

With mknod entries in linux.devices and cgroups entries in linux.resources.devices. Background discussion in [1]. For specifying device cgroups independent of device creation. This makes it easy to distinguish between configs that call for cgroup adjustments (which have linux.resources entries) from those that don't. Without this split, folks interested in making that distinction would have to parse the device section to determine if it included cgroup changes. This will also make it easy to drop either portion (mknod [2] or cgroups [3]) independently of the other if the project decides to do so. Using seperate sections for mknod and cgroups also allows us to avoid the complicated validation rules needed for the combined format mknod/cgroup [4]. Now that there is a section specific to supplying devices, I shifted the default device listing over from config-linux [5]. The /dev/ptmx entry is a bit awkward, since it's not a device, but it seemed to fit better over here. But I would also be fine leaving it with the other mounts in config-linux. The reference links are sorted into two blocks, with kernel-doc links sorted alphabetically followed by man pages sorted alphabetically by section. [1]: https://groups.google.com/a/opencontainers.org/forum/#!topic/dev/y_Fsa2_jJaM Subject: Separate config entries for device mknod and cgroups? Date: Mon, 5 Oct 2015 12:46:55 -0700 Message-ID: <20151005194655.GN28418@odin.tremily.us> [2]: opencontainers#98 [3]: https://groups.google.com/a/opencontainers.org/forum/#!topic/dev/qWHoKs8Fsrk Subject: removal of cgroups from the OCI Linux spec Date: Wed, 28 Oct 2015 17:01:59 +0000 Message-ID: <CAD2oYtO1RMCcUp52w-xXemzDTs+J6t4hS5Mm4mX+uBnVONGDfA@mail.gmail.com> [4]: opencontainers#101 [5]: opencontainers#171 (comment) Signed-off-by: W. Trevor King <wking@tremily.us>

With mknod entries in linux.devices and cgroups entries in linux.resources.devices. Background discussion in [1]. For specifying device cgroups independent of device creation. This makes it easy to distinguish between configs that call for cgroup adjustments (which have linux.resources entries) from those that don't. Without this split, folks interested in making that distinction would have to parse the device section to determine if it included cgroup changes. This will also make it easy to drop either portion (mknod [2] or cgroups [3]) independently of the other if the project decides to do so. Using seperate sections for mknod and cgroups also allows us to avoid the complicated validation rules needed for the combined format mknod/cgroup [4]. Now that there is a section specific to supplying devices, I shifted the default device listing over from config-linux [5]. The /dev/ptmx entry is a bit awkward, since it's not a device, but it seemed to fit better over here. But I would also be fine leaving it with the other mounts in config-linux. The reference links are sorted into two blocks, with kernel-doc links sorted alphabetically followed by man pages sorted alphabetically by section. The cgroup link is new since 2016-01-13 [6]. [1]: https://groups.google.com/a/opencontainers.org/forum/#!topic/dev/y_Fsa2_jJaM Subject: Separate config entries for device mknod and cgroups? Date: Mon, 5 Oct 2015 12:46:55 -0700 Message-ID: <20151005194655.GN28418@odin.tremily.us> [2]: opencontainers#98 [3]: https://groups.google.com/a/opencontainers.org/forum/#!topic/dev/qWHoKs8Fsrk Subject: removal of cgroups from the OCI Linux spec Date: Wed, 28 Oct 2015 17:01:59 +0000 Message-ID: <CAD2oYtO1RMCcUp52w-xXemzDTs+J6t4hS5Mm4mX+uBnVONGDfA@mail.gmail.com> [4]: opencontainers#101 [5]: opencontainers#171 (comment) [6]: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=34a9304a96d6351c2d35dcdc9293258378fc0bd8 Signed-off-by: W. Trevor King <wking@tremily.us>

With mknod entries in linux.devices and cgroups entries in linux.resources.devices. Background discussion in [1]. For specifying device cgroups independent of device creation. This makes it easy to distinguish between configs that call for cgroup adjustments (which have linux.resources entries) from those that don't. Without this split, folks interested in making that distinction would have to parse the device section to determine if it included cgroup changes. This will also make it easy to drop either portion (mknod [2] or cgroups [3]) independently of the other if the project decides to do so. Using seperate sections for mknod and cgroups also allows us to avoid the complicated validation rules needed for the combined format mknod/cgroup [4]. Now that there is a section specific to supplying devices, I shifted the default device listing over from config-linux [5]. The /dev/ptmx entry is a bit awkward, since it's not a device, but it seemed to fit better over here. But I would also be fine leaving it with the other mounts in config-linux. fileMode, uid, and gid are optional, because mknod(2) doesn't need them and specifies the handling when they aren't set [6,7]. Similarly, major/minor numbers are only required for S_IFCHR and S_IFBLK [6]. I've left off wording about required runtime behavior for unset values, because I'd rather address that with a blanket rule [8]. For the cgroup, access is optional because the kernel docs show an example that doesn't write an access field to the devices.deny file [9]. The current kernel docs don't go into much detail on this behavior (I expect unset and 'rwm' are equivalent), but if the kernel doesn't need a value written, the spec should get out of the way and allow users to not specify a value. The reference links are sorted into two blocks, with kernel-doc links sorted alphabetically followed by man pages sorted alphabetically by section. The cgroup link is new since 2016-01-13 [10]. [1]: https://groups.google.com/a/opencontainers.org/forum/#!topic/dev/y_Fsa2_jJaM Subject: Separate config entries for device mknod and cgroups? Date: Mon, 5 Oct 2015 12:46:55 -0700 Message-ID: <20151005194655.GN28418@odin.tremily.us> [2]: opencontainers#98 [3]: https://groups.google.com/a/opencontainers.org/forum/#!topic/dev/qWHoKs8Fsrk Subject: removal of cgroups from the OCI Linux spec Date: Wed, 28 Oct 2015 17:01:59 +0000 Message-ID: <CAD2oYtO1RMCcUp52w-xXemzDTs+J6t4hS5Mm4mX+uBnVONGDfA@mail.gmail.com> [4]: opencontainers#101 [5]: opencontainers#171 (comment) [6]: http://man7.org/linux/man-pages/man2/mknod.2.html#DESCRIPTION [7]: https://github.com/opencontainers/specs/pull/298/files#r51053835 [8]: opencontainers#285 (comment) [9]: https://kernel.org/doc/Documentation/cgroup-v1/devices.txt [10]: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=34a9304a96d6351c2d35dcdc9293258378fc0bd8 Signed-off-by: W. Trevor King <wking@tremily.us>

wking reviewed Sep 10, 2015
View reviewed changes

gao-feng force-pushed the update-runtime-config-linux-md branch from 5d43705 to 5ab8432 Compare September 11, 2015 12:08

wking reviewed Sep 11, 2015
View reviewed changes

gao-feng force-pushed the update-runtime-config-linux-md branch 2 times, most recently from 4dc623c to 107a454 Compare September 14, 2015 02:46

mrunalp reviewed Oct 5, 2015
View reviewed changes

gao-feng force-pushed the update-runtime-config-linux-md branch 3 times, most recently from 01528b3 to 8e1adef Compare October 9, 2015 04:09

wking mentioned this pull request Dec 28, 2015

config: Single, unified config file #284

Merged

gao-feng added 2 commits January 5, 2016 14:13

runtime-config: add example label before json exmaple

0c9a978

Signed-off-by: Gao feng <omarapazanadi@gmail.com>

move the description of user ns mapping to proper file

053f059

They should stay in runtime not config. Signed-off-by: Gao feng <omarapazanadi@gmail.com>

gao-feng force-pushed the update-runtime-config-linux-md branch from 8e1adef to 053f059 Compare January 5, 2016 06:20

hqhq added a commit that referenced this pull request Jan 5, 2016

Merge pull request #171 from gao-feng/update-runtime-config-linux-md

4916d73

move the description of user ns mapping and default files to proper file

hqhq merged commit 4916d73 into opencontainers:master Jan 5, 2016

wking mentioned this pull request Jan 12, 2016

runtime-config-linux: Separate mknod from cgroups #298

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

move the description of user ns mapping and default files to proper file #171

move the description of user ns mapping and default files to proper file #171

gao-feng commented Sep 10, 2015

wking Sep 10, 2015

gao-feng Sep 11, 2015

wking Sep 11, 2015

gao-feng Sep 11, 2015

wking commented Sep 10, 2015

gao-feng commented Sep 11, 2015

wking Sep 11, 2015

gao-feng Sep 14, 2015

wking Sep 14, 2015

gao-feng Sep 14, 2015

wking Sep 14, 2015 via email

mrunalp Oct 5, 2015

wking Oct 5, 2015

mrunalp commented Oct 5, 2015

gao-feng commented Oct 9, 2015

vishh commented Dec 16, 2015

vbatts commented Dec 16, 2015

gao-feng commented Jan 5, 2016

hqhq commented Jan 5, 2016

move the description of user ns mapping and default files to proper file #171

move the description of user ns mapping and default files to proper file #171

Conversation

gao-feng commented Sep 10, 2015

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wking commented Sep 10, 2015

gao-feng commented Sep 11, 2015

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

wking Sep 14, 2015 via email

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mrunalp commented Oct 5, 2015

gao-feng commented Oct 9, 2015

vishh commented Dec 16, 2015

vbatts commented Dec 16, 2015

gao-feng commented Jan 5, 2016

hqhq commented Jan 5, 2016