Re: [PATCH 6/7] mm: vmscan: move dirty pages out of the way until they're flushed

2017-02-03 Thread Michal Hocko
On Fri 03-02-17 15:42:55, Hillf Danton wrote: > > On February 03, 2017 3:20 AM Johannes Weiner wrote: > > @@ -1063,7 +1063,7 @@ static unsigned long shrink_page_list(struct > > list_head *page_list, > > PageReclaim(page) && > > test_bit(PGDAT_WRITE

Re: [PATCH 6/7] mm: vmscan: move dirty pages out of the way until they're flushed

2017-02-02 Thread Hillf Danton
On February 03, 2017 3:20 AM Johannes Weiner wrote: > @@ -1063,7 +1063,7 @@ static unsigned long shrink_page_list(struct list_head > *page_list, > PageReclaim(page) && > test_bit(PGDAT_WRITEBACK, &pgdat->flags)) { >

[PATCH 6/7] mm: vmscan: move dirty pages out of the way until they're flushed

2017-02-02 Thread Johannes Weiner
We noticed a performance regression when moving hadoop workloads from 3.10 kernels to 4.0 and 4.6. This is accompanied by increased pageout activity initiated by kswapd as well as frequent bursts of allocation stalls and direct reclaim scans. Even lowering the dirty ratios to the equivalent of le