Re: [PATCH] meminfo: show /proc/meminfo base on container's memcg

[KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxx>]

On Thu, May 31, 2012 at 2:56 AM, David Rientjes <rientjes@xxxxxxxxxx> wrote:
> On Thu, 31 May 2012, KOSAKI Motohiro wrote:
>
>> > This is tangent to the discussion, we need to revisit why an application
>> > other than a daemon managing a set of memcgs would ever need to know the
>> > information in /proc/meminfo.  No use-case was ever presented in the
>> > changelog and its not clear how this is at all relevant.  So before
>> > changing the kernel, please describe how this actually matters in a real-
>> > world scenario.
>>
>> Huh? Don't you know a meanings of a namespace ISOLATION? isolation mean,
>> isolated container shouldn't be able to access global information. If you
>> want to lean container/namespace concept, tasting openvz or solaris container
>> is a good start.
>
> As I said, LXC and namespace isolation is a tangent to the discussion of
> faking the /proc/meminfo for the memcg context of a thread.

Because of, /proc/meminfo affect a lot of libraries behavior. So, it's not only
application issue. If you can't rewrite _all_ of userland assets, fake meminfo
can't be escaped. Again see alternative container implementation.


>
>> But anyway, I dislike current implementaion. So, I NAK this patch too.
>>
>
> I'm glad you reached that conclusion, but I think you did so for a much
> different (although unspecified) reason.
>
> Thanks.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx.  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