(2012/05/31 9:44), David Rientjes wrote:

On Thu, 31 May 2012, Kamezawa Hiroyuki wrote:

quoted

My test with sysfs node's meminfo seems to work...

[root@rx100-1 qqm]# mount --bind /sys/devices/system/node/node0/meminfo
/proc/meminfo
[root@rx100-1 qqm]# cat /proc/meminfo

Node 0 MemTotal:        8379636 kB

This doesn't seem like a good idea unless the application supports the
"Node 0" prefix in /proc/meminfo.

Of course, /cgroup/memory/..../memory.meminfo , a new file, will use the same
format of /proc/meminfo. Above is just an example of bind-mount.

If any application really cares about the amount of memory available to
it, it should be taught to be memcg aware.  Then do something like

Considering the container as a kind of virtualization, Some kind of
transparent way is required.

Thanks,
-Kame


--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>