Messages in this thread | | | From | Dexuan Cui <> | Subject | RE: Random high CPU utilization in blk-mq with the none scheduler | Date | Tue, 14 Dec 2021 03:09:32 +0000 |
| |
> From: Ming Lei <ming.lei@redhat.com> > Sent: Monday, December 13, 2021 4:53 PM > ... > The single hw queue may be the root cause of your issue, and there > is only single run_work, which can be touched by all CPUs(~200) almost, so > cache > ping-pong could be very serious. > > Jens patch may improve it more or less, please test it. > > Thanks, > Ming
I replied to Jens's email with the test result before your email. :-)
Jen's patch does help in the absence of dc5fc361d891 ("block: attempt direct issue of plug list").
BTW, the throughput in the case of dm_mod.dm_mq_queue_depth=256 is slightly better than the case of dm_mod.dm_mq_queue_depth=2048 with Jen's patch or dc5fc361d891.
| |