From ArchWiki
Revision as of 22:28, 24 September 2010 by Intgr (talk | contribs)
Jump to: navigation, search

cgroups (aka control groups) is a Linux kernel feature to limit, police and account the resource usage of certain processes (actually process groups). For example, cgroups are used by the Linux Containers virtualization solution for resource limiting. Compared to other approaches like the 'nice' command or /etc/security/limits.conf, cgroups are infinitely more flexible.

Control groups can be used in multiple ways:

  • create and manage them on the fly using tools like cgcreate, cgexec, cgclassify etc
  • configure certain users/groups/commands to always run in a group (/etc/cgrules.conf)
  • through other software such as Linux Containers

Unfortunately this feature is often underappreciated due to lack of easy "how-to" style documentation. This is an attempt of fixing the problem. :)

Note: I'm only learning to use cgroups as I type this, so don't take everything I say here as pure gold.


First, install the utilities for managing cgroups; you need the libcgroup package from AUR. makepkg and install it as usual.

Next, you need to define where to mount the cgroup controller virtual file systems. Let's start with the 'cpu' and 'memory' controllers:

Template:File Then run the following to create these directories and mount the controller file systems:

/etc/rc.d/cgconfig start

Now when you list the controller directory, you should see some files (cgroup tunables) in it:

ls -l /mnt/cgroups/memory

You might also want to run this at boot:


Simple usage

Ad-hoc groups

One of the powers of croups is that you can create "ad-hoc" groups on the fly. In fact, you can even grant the privileges to create custom groups to regular users. Run this as root (replace $USER with your user name):

sudo cgcreate -a $USER -g memory,cpu:me

That's it! Now all the tunables in the group "me" are writable by your user:

$ ls -l /mnt/cgroups/memory/me
total 0
-rwxrwxr-x 1 user root 0 Sep 25 00:39 cgroup.event_control
-rwxrwxr-x 1 user root 0 Sep 25 00:39 cgroup.procs
-rwxrwxr-x 1 user root 0 Sep 25 00:39 cpu.rt_period_us
-rwxrwxr-x 1 user root 0 Sep 25 00:39 cpu.rt_runtime_us
-rwxrwxr-x 1 user root 0 Sep 25 00:39 cpu.shares
-rwxrwxr-x 1 user root 0 Sep 25 00:39 notify_on_release
-rwxrwxr-x 1 user root 0 Sep 25 00:39 tasks

Let's say that, as a normal user, you want to run a bash shell under a new sub-group called 'foo'. That's simple:

cgcreate -g memory,cpu:me/foo
cgexec -g memory,cpu:me/foo bash

There we go! Just to make sure:

$ cat /proc/self/cgroup

A new subdirectory was greated for this group. To limit the memory usage of all processes in this group to 10 MB, run the following:

$ echo 10000000 > /mnt/cgroups/memory/me/foo/memory.limit_in_bytes

Note that the memory limit applies to RAM use only -- once tasks hit this limit, they will begin to swap. But it won't affect the performance of other processes significantly.

Similarly you can change the CPU priority ("shares") of this group. By default all groups have 1024 shares. A group with 100 shares will get a ~10% portion of the CPU time:

$ echo 100 > /mnt/cgroups/cpu/me/foo/cpu.shares

You can find more tunables or statistics by listing the cgroup directory.

You can also change the cgroup of already running processes. To move all 'bash' commands to this group:

$ pidof bash
13244 13266
$ cgclassify -g memory,cpu:me/foo `pidof bash`
$ cat /proc/13244/cgroup

Persistent group configuration

If you want your cgroups to be created at boot, you can define them in the configuration file instead. For example, the "me" group and "me/foo" definitions would look like this: