From: Jens Axboe <axboe@fb.com> To: Linus Torvalds <torvalds@linux-foundation.org>, Marc MERLIN <marc@merlins.org>, Kent Overstreet <kent.overstreet@gmail.com>, Tejun Heo <tj@kernel.org> Cc: Michal Hocko <mhocko@kernel.org>, Vlastimil Babka <vbabka@suse.cz>, linux-mm <linux-mm@kvack.org>, LKML <linux-kernel@vger.kernel.org>, "Joonsoo Kim" <iamjoonsoo.kim@lge.com>, Greg Kroah-Hartman <gregkh@linuxfoundation.org> Subject: Re: 4.8.8 kernel trigger OOM killer repeatedly when I have lots of RAM that should be free Date: Wed, 30 Nov 2016 11:27:43 -0700 [thread overview] Message-ID: <0c7df460-90a3-b71e-3965-abda00336ac9@fb.com> (raw) In-Reply-To: <CA+55aFzPQpvttSryRL3+EWeY7X+uFWOk2V+mM8JYm7ba+X1gHg@mail.gmail.com> On 11/30/2016 11:14 AM, Linus Torvalds wrote: > On Wed, Nov 30, 2016 at 9:47 AM, Marc MERLIN <marc@merlins.org> wrote: >> >> I gave it a thought again, I think it is exactly the nasty situation you >> described. >> bcache takes I/O quickly while sending to SSD cache. SSD fills up, now >> bcache can't handle IO as quickly and has to hang until the SSD has been >> flushed to spinning rust drives. >> This actually is exactly the same as filling up the cache on a USB key >> and now you're waiting for slow writes to flash, is it not? > > It does sound like you might hit exactly the same kind of situation, yes. > > And the fact that you have dmcrypt running too just makes things pile > up more. All those IO's end up slowed down by the scheduling too. > > Anyway, none of this seems new per se. I'm adding Kent and Jens to the > cc (Tejun already was), in the hope that maybe they have some idea how > to control the nasty worst-case behavior wrt workqueue lockup (it's > not really a "lockup", it looks like it's just hundreds of workqueues > all waiting for IO to complete and much too deep IO queues). Honestly, the easiest would be to wire it up to the blk-wbt stuff that is queued up for 4.10, which attempts to limit the queue depths to something reasonable instead of letting them run amok. This is largely (exclusively, almost) a problem with buffered writeback. On devices utilizing the stacked interface, they never get any depth throttling. Obviously it's worse if each IO ends up queueing work, but it's a big problem even if they do not. > I think it's the traditional "throughput is much easier to measure and > improve" situation, where making queues big help some throughput > situation, but ends up causing chaos when things go south. Yes, and the longer queues never buy you anything, but they end up causing tons of problems at the other end of the spectrum. Still makes sense to limit dirty memory for highmem, though. -- Jens Axboe
WARNING: multiple messages have this Message-ID (diff)
From: Jens Axboe <axboe@fb.com> To: Linus Torvalds <torvalds@linux-foundation.org>, Marc MERLIN <marc@merlins.org>, Kent Overstreet <kent.overstreet@gmail.com>, Tejun Heo <tj@kernel.org> Cc: Michal Hocko <mhocko@kernel.org>, Vlastimil Babka <vbabka@suse.cz>, linux-mm <linux-mm@kvack.org>, LKML <linux-kernel@vger.kernel.org>, Joonsoo Kim <iamjoonsoo.kim@lge.com>, Greg Kroah-Hartman <gregkh@linuxfoundation.org> Subject: Re: 4.8.8 kernel trigger OOM killer repeatedly when I have lots of RAM that should be free Date: Wed, 30 Nov 2016 11:27:43 -0700 [thread overview] Message-ID: <0c7df460-90a3-b71e-3965-abda00336ac9@fb.com> (raw) In-Reply-To: <CA+55aFzPQpvttSryRL3+EWeY7X+uFWOk2V+mM8JYm7ba+X1gHg@mail.gmail.com> On 11/30/2016 11:14 AM, Linus Torvalds wrote: > On Wed, Nov 30, 2016 at 9:47 AM, Marc MERLIN <marc@merlins.org> wrote: >> >> I gave it a thought again, I think it is exactly the nasty situation you >> described. >> bcache takes I/O quickly while sending to SSD cache. SSD fills up, now >> bcache can't handle IO as quickly and has to hang until the SSD has been >> flushed to spinning rust drives. >> This actually is exactly the same as filling up the cache on a USB key >> and now you're waiting for slow writes to flash, is it not? > > It does sound like you might hit exactly the same kind of situation, yes. > > And the fact that you have dmcrypt running too just makes things pile > up more. All those IO's end up slowed down by the scheduling too. > > Anyway, none of this seems new per se. I'm adding Kent and Jens to the > cc (Tejun already was), in the hope that maybe they have some idea how > to control the nasty worst-case behavior wrt workqueue lockup (it's > not really a "lockup", it looks like it's just hundreds of workqueues > all waiting for IO to complete and much too deep IO queues). Honestly, the easiest would be to wire it up to the blk-wbt stuff that is queued up for 4.10, which attempts to limit the queue depths to something reasonable instead of letting them run amok. This is largely (exclusively, almost) a problem with buffered writeback. On devices utilizing the stacked interface, they never get any depth throttling. Obviously it's worse if each IO ends up queueing work, but it's a big problem even if they do not. > I think it's the traditional "throughput is much easier to measure and > improve" situation, where making queues big help some throughput > situation, but ends up causing chaos when things go south. Yes, and the longer queues never buy you anything, but they end up causing tons of problems at the other end of the spectrum. Still makes sense to limit dirty memory for highmem, though. -- Jens Axboe -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2016-11-30 18:28 UTC|newest] Thread overview: 113+ messages / expand[flat|nested] mbox.gz Atom feed top 2016-11-21 15:43 4.8.8 kernel trigger OOM killer repeatedly when I have lots of RAM that should be free Marc MERLIN 2016-11-21 16:30 ` Marc MERLIN 2016-11-21 21:50 ` Vlastimil Babka 2016-11-21 21:50 ` Vlastimil Babka 2016-11-21 21:56 ` Marc MERLIN 2016-11-21 21:56 ` Marc MERLIN 2016-11-21 23:03 ` [PATCH] block,blkcg: use __GFP_NOWARN for best-effort allocations in blkcg Tejun Heo 2016-11-21 23:03 ` Tejun Heo 2016-11-22 15:47 ` Vlastimil Babka 2016-11-22 15:47 ` Vlastimil Babka 2016-11-22 16:48 ` Tejun Heo 2016-11-22 16:48 ` Tejun Heo 2016-11-22 22:13 ` Linus Torvalds 2016-11-22 22:13 ` Linus Torvalds 2016-11-23 8:50 ` Vlastimil Babka 2016-11-23 8:50 ` Vlastimil Babka 2016-11-28 17:19 ` Tejun Heo 2016-11-28 17:19 ` Tejun Heo 2016-11-29 7:25 ` Michal Hocko 2016-11-29 7:25 ` Michal Hocko 2016-11-29 16:38 ` Tejun Heo 2016-11-29 16:38 ` Tejun Heo 2016-11-29 16:57 ` Vlastimil Babka 2016-11-29 16:57 ` Vlastimil Babka 2016-11-29 17:13 ` Michal Hocko 2016-11-29 17:13 ` Michal Hocko 2016-11-29 17:17 ` Linus Torvalds 2016-11-29 17:17 ` Linus Torvalds 2016-11-29 17:28 ` Michal Hocko 2016-11-29 17:28 ` Michal Hocko 2016-11-29 17:48 ` Linus Torvalds 2016-11-29 17:48 ` Linus Torvalds 2016-11-22 16:00 ` Jens Axboe 2016-11-22 16:00 ` Jens Axboe 2016-11-22 16:06 ` 4.8.8 kernel trigger OOM killer repeatedly when I have lots of RAM that should be free Marc MERLIN 2016-11-22 16:06 ` Marc MERLIN 2016-11-22 16:14 ` Vlastimil Babka 2016-11-22 16:14 ` Vlastimil Babka 2016-11-22 16:25 ` Michal Hocko 2016-11-22 16:25 ` Michal Hocko 2016-11-22 16:47 ` Marc MERLIN 2016-11-22 16:47 ` Marc MERLIN 2016-11-22 16:38 ` Greg Kroah-Hartman 2016-11-22 16:38 ` Greg Kroah-Hartman 2016-11-29 16:25 ` Michal Hocko 2016-11-29 16:25 ` Michal Hocko 2016-11-29 16:25 ` Michal Hocko 2016-11-29 16:43 ` Patch "mm, oom: stop pre-mature high-order OOM killer invocations" has been added to the 4.8-stable tree gregkh 2016-11-29 16:43 ` 4.8.8 kernel trigger OOM killer repeatedly when I have lots of RAM that should be free Greg Kroah-Hartman 2016-11-29 16:43 ` Greg Kroah-Hartman 2016-11-22 19:38 ` Linus Torvalds 2016-11-22 19:38 ` Linus Torvalds 2016-11-23 6:34 ` Michal Hocko 2016-11-23 6:34 ` Michal Hocko 2016-11-23 6:53 ` Hillf Danton 2016-11-23 6:53 ` Hillf Danton 2016-11-23 7:00 ` Michal Hocko 2016-11-23 7:00 ` Michal Hocko 2016-11-23 9:18 ` Vlastimil Babka 2016-11-23 9:18 ` Vlastimil Babka 2016-11-28 7:23 ` Michal Hocko 2016-11-28 7:23 ` Michal Hocko 2016-11-28 20:55 ` Marc MERLIN 2016-11-29 15:55 ` Marc MERLIN 2016-11-29 15:55 ` Marc MERLIN 2016-11-29 16:07 ` Michal Hocko 2016-11-29 16:07 ` Michal Hocko 2016-11-29 16:34 ` Marc MERLIN 2016-11-29 16:34 ` Marc MERLIN 2016-11-29 17:07 ` Linus Torvalds 2016-11-29 17:07 ` Linus Torvalds 2016-11-29 17:40 ` Marc MERLIN 2016-11-29 17:40 ` Marc MERLIN 2016-11-29 18:01 ` Linus Torvalds 2016-11-29 18:01 ` Linus Torvalds 2016-11-30 17:47 ` Marc MERLIN 2016-11-30 17:47 ` Marc MERLIN 2016-11-30 18:14 ` Linus Torvalds 2016-11-30 18:21 ` Marc MERLIN 2016-11-30 18:21 ` Marc MERLIN 2016-11-30 18:27 ` Jens Axboe [this message] 2016-11-30 18:27 ` Jens Axboe 2016-11-30 20:30 ` Tejun Heo 2016-11-30 20:30 ` Tejun Heo 2016-12-01 13:50 ` Kent Overstreet 2016-12-01 13:50 ` Kent Overstreet 2016-12-01 18:16 ` Linus Torvalds 2016-12-01 18:16 ` Linus Torvalds 2016-12-01 18:30 ` Jens Axboe 2016-12-01 18:30 ` Jens Axboe 2016-12-01 18:37 ` Linus Torvalds 2016-12-01 18:37 ` Linus Torvalds 2016-12-01 18:46 ` Jens Axboe 2016-12-01 18:46 ` Jens Axboe 2016-11-29 20:11 ` Holger Hoffstätte 2016-11-29 23:01 ` Marc MERLIN 2016-11-29 23:01 ` Marc MERLIN 2016-11-30 13:58 ` Tetsuo Handa 2016-11-30 13:58 ` Tetsuo Handa 2017-05-02 4:12 ` Marc MERLIN 2017-05-02 4:12 ` Marc MERLIN 2017-05-02 7:44 ` Michal Hocko 2017-05-02 7:44 ` Michal Hocko 2017-05-02 14:15 ` Marc MERLIN 2017-05-02 14:15 ` Marc MERLIN 2017-05-02 10:44 ` Tetsuo Handa 2017-05-02 10:44 ` Tetsuo Handa 2016-11-29 16:15 ` Marc MERLIN 2016-11-29 16:15 ` Marc MERLIN 2016-11-22 21:46 ` Simon Kirby 2016-11-22 21:46 ` Simon Kirby 2016-11-28 8:06 ` Vlastimil Babka 2016-11-28 8:06 ` Vlastimil Babka
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=0c7df460-90a3-b71e-3965-abda00336ac9@fb.com \ --to=axboe@fb.com \ --cc=gregkh@linuxfoundation.org \ --cc=iamjoonsoo.kim@lge.com \ --cc=kent.overstreet@gmail.com \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-mm@kvack.org \ --cc=marc@merlins.org \ --cc=mhocko@kernel.org \ --cc=tj@kernel.org \ --cc=torvalds@linux-foundation.org \ --cc=vbabka@suse.cz \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.