[PATCH V2] Documentation: cgroups: Add notes on removing PIDs from tasks and cgroups requiring setup

Paul Menage menage at google.com
Thu Mar 10 19:00:09 PST 2011


On Thu, Mar 10, 2011 at 9:07 AM, Eric B Munson <emunson at mgebm.net> wrote:
> The cgroup documentation does not specify how a process can be removed from a
> particular group.  This patch adds a note at the end of the simple example
> about how this is done.  Also, some cgroups (like cpusets) require user input
> before a new group can be used.  This is noted in the patch as well.
>
> Signed-off-by: Eric B Munson <emunson at mgebm.net>

Acked-by: Paul Menage <menage at google.com>

> ---
> Changes from V1:
>  Rework the note about removing a process from a cgroup
>  Add note about ns cgroup causing failure when moving processes between groups
>
>  Documentation/cgroups/cgroups.txt |   12 ++++++++++++
>  1 files changed, 12 insertions(+), 0 deletions(-)
>
> diff --git a/Documentation/cgroups/cgroups.txt b/Documentation/cgroups/cgroups.txt
> index 44b8b7a..cbdfb7d 100644
> --- a/Documentation/cgroups/cgroups.txt
> +++ b/Documentation/cgroups/cgroups.txt
> @@ -349,6 +349,10 @@ To mount a cgroup hierarchy with all available subsystems, type:
>  The "xxx" is not interpreted by the cgroup code, but will appear in
>  /proc/mounts so may be any useful identifying string that you like.
>
> +Note: Some subsystems do not work without some user input first.  For instance,
> +if cpusets are enabled the user will have to populate the cpus and mems files
> +for each new cgroup created before that group can be used.
> +
>  To mount a cgroup hierarchy with just the cpuset and memory
>  subsystems, type:
>  # mount -t cgroup -o cpuset,memory hier1 /dev/cgroup
> @@ -426,6 +430,14 @@ You can attach the current shell task by echoing 0:
>
>  # echo 0 > tasks
>
> +Note: Since every task is always a member of exactly one cgroup in each
> +mounted hierarchy, to remove a task from its current cgroup you must
> +move it into a new cgroup (possibly the root cgroup) by writing to the
> +new cgroup's tasks file.
> +
> +Note: If the ns cgroup is active, moving a process to another cgroup can
> +fail.
> +
>  2.3 Mounting hierarchies by name
>  --------------------------------
>
> --
> 1.7.1
>
>


More information about the Containers mailing list