在 2019/8/22 上午2:00, Daniel Jordan 写道: >> > > This is system-wide right, not per container? Even per container, 89 usec isn't much contention over 20 seconds. You may want to give this a try: yes, perf lock show the host info. > > https://git.kernel.org/pub/scm/linux/kernel/git/wfg/vm-scalability.git/tree/case-lru-file-readtwice> > It's also synthetic but it stresses lru_lock more than just anon alloc/free. It hits the page activate path, which is where we see this lock in our database, and if enough memory is configured lru_lock also gets stressed during reclaim, similar to [1]. Thanks for the sharing, this patchset can not help the [1] case, since it's just relief the per container lock contention now. Yes, readtwice case could be more sensitive for this lru_lock changes in containers. I may try to use it in container with some tuning. But anyway, aim9 is also pretty good to show the problem and solutions. :) > > It'd be better though, as Michal suggests, to use the real workload that's causing problems. Where are you seeing contention? We repeatly create or delete a lot of different containers according to servers load/usage, so normal workload could cause lots of pages alloc/remove. aim9 could reflect part of scenarios. I don't know the DB scenario yet. > >> With this patch series, lruvec->lru_lock show no contentions >> &(&lruvec->lru_l... 8 0 0 0 0 0 >> >> and aim9 page_test/brk_test performance increased 5%~50%. > > Where does the 50% number come in? The numbers below seem to only show ~4% boost. the Setddev/CoeffVar case has about 50% performance increase. one of container's mmtests result as following: Stddev page_test 245.15 ( 0.00%) 189.29 ( 22.79%) Stddev brk_test 1258.60 ( 0.00%) 629.16 ( 50.01%) CoeffVar page_test 0.71 ( 0.00%) 0.53 ( 26.05%) CoeffVar brk_test 1.32 ( 0.00%) 0.64 ( 51.14%)
