linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [RFC PATCH 0/3] Revisit NUMA imbalance tolerance and fork balancing
@ 2020-11-17 13:42 Mel Gorman
  2020-11-17 13:42 ` [PATCH 1/3] sched/numa: Rename nr_running and break out the magic number Mel Gorman
                   ` (2 more replies)
  0 siblings, 3 replies; 16+ messages in thread
From: Mel Gorman @ 2020-11-17 13:42 UTC (permalink / raw)
  To: LKML
  Cc: Ingo Molnar, Peter Zijlstra, Vincent Guittot, Valentin Schneider,
	Juri Lelli, Mel Gorman

When NUMA and CPU balancing were reconciled, there was an attempt to allow
a degree of imbalance but it caused more problems than it solved. Instead,
imbalance was only allowed with an almost idle NUMA domain. A lot of the
problems have since been addressed so it's time for a revisit. There is
also an issue with how fork is balanced across threads. It's mentioned
in this context as patch 2 and 3 should share similar behaviour in terms
of a nodes utilisation.

Patch 1 is just a cosmetic rename

Patch 2 allows a "floating" imbalance to exist so communicating tasks can
	remain on the same domain until utilisation is higher. It aims
	to balance compute availability with memory bandwidth.

Patch 3 is the interesting one. Currently fork can allow a NUMA node
	to be completely utilised as long as there are idle CPUs until
	the load balancer gets involved. This caused serious problems
	with a real workload that unfortunately I cannot share many
	details about but there is a proxy reproducer.

 kernel/sched/fair.c | 41 ++++++++++++++++++++++++-----------------
 1 file changed, 24 insertions(+), 17 deletions(-)

^ permalink raw reply	[flat|nested] 16+ messages in thread

end of thread, other threads:[~2020-11-22 14:50 UTC | newest]

Thread overview: 16+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-11-17 13:42 [RFC PATCH 0/3] Revisit NUMA imbalance tolerance and fork balancing Mel Gorman
2020-11-17 13:42 ` [PATCH 1/3] sched/numa: Rename nr_running and break out the magic number Mel Gorman
2020-11-17 13:42 ` [PATCH 2/3] sched/numa: Allow a floating imbalance between NUMA nodes Mel Gorman
2020-11-17 14:16   ` Peter Zijlstra
2020-11-17 14:43     ` Mel Gorman
2020-11-17 14:24   ` Vincent Guittot
2020-11-17 14:44     ` Mel Gorman
2020-11-17 13:42 ` [PATCH 3/3] sched/numa: Limit the amount of imbalance that can exist at fork time Mel Gorman
2020-11-17 14:18   ` Peter Zijlstra
2020-11-17 14:31     ` Vincent Guittot
2020-11-17 15:17       ` Mel Gorman
2020-11-17 15:53         ` Vincent Guittot
2020-11-17 17:28           ` Mel Gorman
2020-11-18 16:06             ` Vincent Guittot
2020-11-18 16:50               ` Mel Gorman
2020-11-22 15:04   ` [sched/numa] e7f28850ea: unixbench.score 1.5% improvement kernel test robot

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).