Re: [PATCH 3/3] mm: memcontrol: fix cgroup creation failure after many small jobs

[Johannes Weiner]

On Tue, Jun 21, 2016 at 01:16:51PM +0300, Vladimir Davydov wrote:
> On Fri, Jun 17, 2016 at 12:25:16PM -0400, Johannes Weiner wrote:
> > After this patch, the IDs get released upon cgroup destruction and the
> > cache and css objects get released once memory reclaim kicks in.
> 
> With 65K cgroups it will take the reclaimer a substantial amount of time
> to iterate over all of them, which might result in latency spikes.
> Probably, to avoid that, we could move pages from a dead cgroup's lru to
> its parent's one on offline while still leaving dead cgroups pinned,
> like we do in case of list_lru entries.

Yep, that is true. list_lru is a bit easier because the walker stays
in the context of the original LRU list, whereas the cache/anon LRUs
are not. We'd have to have mem_cgroup_page_lruvec() etc. do a parent
walk to find the closest live ancestor. Maybe you have a better idea?

But it's definitely worth considering. I'll think more about it.

> > Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx>
> 
> Reviewed-by: Vladimir Davydov <vdavydov@xxxxxxxxxxxxx>

Thanks!

> > +static struct idr mem_cgroup_idr;
> 
> static DEFINE_IDR(mem_cgroup_idr);

Oops, good catch. Andrew, could you kindly fold this?