util-vserver:Cgroups - Revision history

Yang: /* Prerequisites */ additional flag required

2012-07-06T13:16:28Z

‎Prerequisites: additional flag required

2012-03-09T09:32:14Z

‎using cgroup to enforce memory limits

2012-03-09T09:31:09Z

‎using cgroup to enforce memory limits

2012-01-22T16:03:20Z

‎Draft - Distributing cpu shares with cgroups

2011-10-21T21:19:52Z

2011-10-10T09:14:22Z

Reverted edits by JBustos36 (Talk) to last revision by 71.11.175.78

2011-10-10T08:59:09Z

2011-07-20T19:35:35Z

‎Requirements

2011-05-14T16:59:09Z

‎Kernel configuration

2011-02-06T23:18:02Z

‎Ben's install on Debian Squeeze/Sid

@@ Line 15: / Line 15: @@
 If you were to mount the <tt>cgroup</tt> Control Groups filesystem manually, you would use something like:
 : <tt># mkdir /dev/cgroup
-: # mount -t cgroup -o ''<subsystems>'' /dev/cgroup</tt>
+: # mount -t cgroup -o ''<subsystems>'' none /dev/cgroup</tt>
 Where <tt>''<subsystems>''</tt> is something like <tt>cpuset,memory</tt>.

← Older revision		Revision as of 09:32, 9 March 2012
Line 129:		Line 129:

	Note: cgroup memory limits are to replace rss.soft and rss.hard some time in the future.		Note: cgroup memory limits are to replace rss.soft and rss.hard some time in the future.
		+
	Note: from kernel 3.2.x+ you HAVE to boot with the kernel parameter swapaccount=1 or swap accounting is disabled		Note: from kernel 3.2.x+ you HAVE to boot with the kernel parameter swapaccount=1 or swap accounting is disabled

← Older revision		Revision as of 09:31, 9 March 2012
Line 129:		Line 129:

	Note: cgroup memory limits are to replace rss.soft and rss.hard some time in the future.		Note: cgroup memory limits are to replace rss.soft and rss.hard some time in the future.
		+	Note: from kernel 3.2.x+ you HAVE to boot with the kernel parameter swapaccount=1 or swap accounting is disabled
		+
		+	Limiting the memory and not the swap mean as soon as the memory limit is reached the guest will swap until it fills out the swap completly so swap limits are somewhat necessary in preventing runaway process
		+
	When you wish the guests to see only their limited memory pool, be sure to include VIRT_MEM in your cflags config file.		When you wish the guests to see only their limited memory pool, be sure to include VIRT_MEM in your cflags config file.

@@ Line 32: / Line 32: @@
 echo '512' > /dev/cgroup/<guest name>/cpu.shares
-The share you get is equal to the guest's share divided by the sum of the cpu shares of all the guest. So for exemple :
+The share you get is equal to the guest's share divided by the sum of the cpu shares of all the guest. So for example :
 <pre>

← Older revision		Revision as of 21:19, 21 October 2011
Line 263:		Line 263:

	Note that the "memory" cgroup subsystem is omitted as Squeeze has the legacy memory controls through rlimits compiled in. It is possible to add "memory" to the cgroup fstab line and use the cgroup based memory control. Please add any success with this to this page.		Note that the "memory" cgroup subsystem is omitted as Squeeze has the legacy memory controls through rlimits compiled in. It is possible to add "memory" to the cgroup fstab line and use the cgroup based memory control. Please add any success with this to this page.
		+
		+	[[Category:Configuration]]

@@ Line 147: / Line 147: @@
 == Ben's install on Debian Lenny ==
-I used the kernels from<span class="plainlinks">[http://thoughtmechanics.com/ <span style="color:black;font-weight:normal;text-decoration:none!important;background:none!important; text-decoration:none;">web site design atlanta</span>]</span> described at<span class="plainlinks">[http://thoughtmechanics.com/ <span style="color:black;font-weight:normal;text-decoration:none!important;background:none!important; text-decoration:none;">web design companies atlanta</span>]</span> I've done this on a few versions, works for 2.6.31.7 with patch vs2.3.0.36.27 on amd64, also 2.6.31.11 with patch vs2.3.0.36.28. I used the stock Lenny util-vserver, patched as described below. The kernel config is critically important, with specific cgroup options necessary in order to get cgroups working in this way. Check the configs for the [http://repo.psand.net] kernels to see which ones I used.
+I used the kernels from [http://repo.psand.net], described at [http://kernels.bristolwireless.net/]. I've done this on a few versions, works for 2.6.31.7 with patch vs2.3.0.36.27 on amd64, also 2.6.31.11 with patch vs2.3.0.36.28. I used the stock Lenny util-vserver, patched as described below. The kernel config is critically important, with specific cgroup options necessary in order to get cgroups working in this way. Check the configs for the [http://repo.psand.net] kernels to see which ones I used.
 ==== Getting Lenny Ready ====

@@ Line 87: / Line 87: @@
 This feature is currently available in <tt>patch-2.6.32.20-vs2.3.0.36.29.6.diff</tt> and is in testing phase as of this patch set so report any bugs to the mailing list.
-To get the hard limit setup on every vServer start you need a recent utils package. It worked for me with: <tt>0.30.216-pre2864</tt>.  (Download from [http://people.linux-vserver.org/~dhozac/t/uv-testing/ util-vserver prereleases])
+To get the hard limit setup on every vServer start you need a recent utils package. It worked for me with: <tt>0.30.216-pre2864</tt>.  (Download from [http://people.linux-vserver.org/~dhozac/t/uv-testing/ util-vserver prereleases]) (also see note at top of page regarding CONFIG_CGROUP_NS, which can usually be found by grepping /proc/config.gz or /boot/config-`uname -r`)
 Before trying to setup limits for one guest you should mount the cgroup filesystem:

@@ Line 3: / Line 3: @@
 == Kernel configuration ==
-When configuring your kernel for cgroups with util-vserver you must make sure <tt>CONFIG_CGROUP_NS</tt> ('''CGroup Namespaces''') is unset for the time being.
+When configuring your kernel for cgroups with util-vserver you must make sure <tt>CONFIG_CGROUP_NS</tt> ('''CGroup Namespaces''') is unset with util-vserver version lower than 0.30.216-pre2882.
 '''CGroup Namespaces''' are a different approach to namespaces than that used by Linux vServer, and are not currently supported.

@@ Line 250: / Line 250: @@
 == Ben's install on Debian Squeeze/Sid ==
-Squeeze is due to ship with the 2.6.32 kernel. Currently the package linux-image-2.6.32-5-vserver-amd64 works well for cgroup scheduling. The following steps are simplest way to set it up:
+These instructions are for Debian's own packages.
 * mkdir /etc/vservers/.defaults/cgroup
@@ Line 258: / Line 260: @@
 * reboot the server
-Instructions for setting particular parameters are the same as for Lenny.
+Instructions for setting particular parameters are the same as for Lenny. The reason for specifying the cgroup subsystems is that if the namespace subsystem "ns" is included, Linux-Vserver will not work. The /etc/fstab line above mounts /dev/cgroup with all the available subsystems excluding "ns".
-The reason for specifying the cgroup subsystems is that if the namespace subsystem "ns" is included, Linux-Vserver will not work. The /etc/fstab line above mounts /dev/cgroup with all the available subsystems excluding "ns".
+Note that the "memory" cgroup subsystem is omitted as Squeeze has the legacy memory controls through rlimits compiled in. It is possible to add "memory" to the cgroup fstab line and use the cgroup based memory control. Please add any success with this to this page.