On Wed 26-02-20 09:00:57, Shakeel Butt wrote: > On Wed, Feb 26, 2020 at 1:08 AM Michal Hocko <mhocko@xxxxxxxxxx> wrote: > > > > On Tue 25-02-20 14:30:03, Shakeel Butt wrote: > > > On Tue, Feb 25, 2020 at 1:10 AM Michal Hocko <mhocko@xxxxxxxxxx> wrote: > > > > > > > [snip] > > > > > > > > The proper fix should, however, check the amount of reclaimable pages > > > > and back off if they cannot meet the target IMO. We cannot rely on the > > > > general reclaimability here because that could really be thrashing. > > > > > > > > > > "check the amount of reclaimable pages" vs "cannot rely on the general > > > reclaimability"? Can you clarify? > > > > kswapd targets the high watermark and if your reclaimable memory (aka > > zone_reclaimable_pages) is lower than the high wmark then it cannot > > simply satisfy that target, right? Keeping reclaim in that situations > > seems counter productive to me because you keep evicting pages that > > might be reused without any feedback mechanism on the actual usage. > > Please see my other reply. > > > > I understand and agree with the argument that if reclaimable pages are > less than high wmark then no need to reclaim. Regarding not depending > on general reclaimability, I thought you meant that even if > reclaimable pages are over high wmark, we might not want to continue > the reclaim to not cause thrashing. Is that right? That is a completely different story. I would stick with the pathological problem reported here. General threshing problem is much more complex and harder to provide a solution for without introducing a lot of policy into the reclaim. -- Michal Hocko SUSE Labs
