Date: Sat, 23 Apr 2022 09:35:49 -0700 (PDT) From: "Jeffrey Bouquet" <jbtakk@iherebuywisely.com> To: "Pete Wright" <pete@nomadlogic.org> Cc: "freebsd-current" <freebsd-current@freebsd.org> Subject: Re: Chasing OOM Issues - good sysctl metrics to use? Message-ID: <E1niIjd-0000mL-4y@rmmprod06.runbox> In-Reply-To: <d7f11201-bab8-ecec-2b34-44baa162dc51@nomadlogic.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 22 Apr 2022 16:46:20 -0700, Pete Wright <pete@nomadlogic.org> wrote: >=20 >=20 > On 4/22/22 13:39, tech-lists wrote: > > Hi, > > > > On Thu, Apr 21, 2022 at 07:16:42PM -0700, Pete Wright wrote: > >> hello - > >> > >> on my workstation running CURRENT (amd64/32g of ram) i've been running > >> into a scenario where after 4 or 5 days of daily use I get an OOM event > >> and both chromium and firefox are killed.=C2=A0 then in the next day o= r so > >> the system will become very unresponsive in the morning when i unlock = my > >> screensaver in the morning forcing a manual power cycle. > > > > I have the following set in /etc/sysctl.conf on a stable/13=20 > > workstation. Am using zfs with 32GB RAM. > > > > vm.pageout_oom_seq=3D120 > > vm.pfault_oom_attempts=3D-1 > > vm.pageout_update_period=3D0 > > > > Since setting these here, OOM is a rarity. I don't profess to exactly=20 > > know > > what they do in detail though. But my experience since these were set > > is hardly any OOM and big users of memory like firefox don't crash. >=20 > nice, i will give those a test next time i crash which will be by next=20 > thurs if the pattern continues. >=20 > looking at the sysctl descriptions: > vm.pageout_oom_seq: back-to-back calls to oom detector to start OOM > vm.pfault_oom_attempts: Number of page allocation attempts in page fault= =20 > handler before it triggers OOM handling > vm.pageout_update_period: Maximum active LRU update period >=20 > i could certainly see how those could be helpful.=C2=A0 in an ideal world= i'd=20 > find the root cause of the system lock-ups, but it would be nice to just= =20 > move on from this :) >=20 > cheers, > -p >=20 > --=20 > Pete Wright > pete@nomadlogic.org > @nomadlogicLA I don't know if it would help this discussion, but I installed security/pic= ocrypt and memory usage started going thru the roof*. Maybe that port could be use= d to test any fix found in this thread, iow if that port runs better without /wi= th some sysctl... if could be used for the issues discussed here so far.=20=20 * and I had to desintall. FWIW I had set the three vm listed above a sh= ort while prior.=20=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?E1niIjd-0000mL-4y>