**** BEGIN LOGGING AT Fri Jun 05 02:59:57 2020
Jun 05 07:05:49 <Tofe> o/
Jun 05 07:06:19 <Herrie> Morning
Jun 05 07:16:09 <Herrie> JaMa: These are the results: https://paste.ubuntu.com/p/ZK3xQT29hH/
Jun 05 07:16:20 <Herrie> Seems to be within minutes of previously and this is without tmpfs even
Jun 05 07:18:32 <Tofe> When I'll leaving for a week of vacations, I'll be able to start this kind of script :p
Jun 05 07:18:40 <Tofe> +be
Jun 05 07:20:52 <Herrie> Tofe: Yeah LOL
Jun 05 07:21:02 <Herrie> JaMa: Chromium failed with this error: https://paste.ubuntu.com/p/mrVChyWC4h/
Jun 05 07:21:15 <Herrie> Seems it segfaulted, no ICE. Not sure if that's memory related or could be something else?
Jun 05 07:21:30 <Herrie> With previous memory issues I would mostly get ICE errors it seems
Jun 05 07:26:54 <Herrie> JaMa: What would be a good size for tmpfs?
Jun 05 07:34:24 <Herrie> Though based on the results it seems tmpfs has minimal impact on builds? What I can see we're talking about 1 or 2 minutes in most cases? At least based on some of the results in your tables
Jun 05 10:11:38 <JaMa> Herrie: wasn't it OOMK?
Jun 05 10:12:18 <JaMa> Herrie: yes tmpfs has very little impact, check what I wrote 2-3 days ago
Jun 05 10:13:36 <JaMa> You can compare tmpfs speed on something like chromium (with the individual build script), builting it all won't fit in our ram
Jun 05 10:33:19 <Herrie> JaMa: How to tell if it was OOMK?
Jun 05 11:01:33 <JaMa> "dmesg | grep Out" if it's long enough otherwise kernel logs in /var/log
Jun 05 11:02:08 <JaMa> in 5.4 kernel it looks like this:
Jun 05 11:02:16 <JaMa> Jun  1 19:59:24 jama kernel: [40231.118662] Out of memory: Killed process 3682894 (cc1plus) total-vm:2180424kB, anon-rss:2046756kB, file-rss:0kB, shmem-rss:0kB, UID:1000 pgtables:4168kB oom_score_adj:0
Jun 05 11:02:30 <JaMa> in older kernel the format was IIRC slightly different, but it contained "Out" as well
Jun 05 11:02:44 <JaMa> /var/log/kern.log
Jun 05 11:20:26 <JaMa> we were almost done with the builds on jenkins, so I've added few small fixes to sleepd and powerd and merged herrie/acg branch here and included small zeus upgrade in next testing builds
Jun 05 11:24:13 <Tofe> JaMa: great, thanks
Jun 05 11:26:10 <JaMa> sarcasm from you almost catching up with local builds? :)
Jun 05 11:26:48 <JaMa> I fear new gstreamer will cause another qwe rebuild, sorry.
Jun 05 11:34:11 <Tofe> JaMa: no worries, I choose when to pull these changes
Jun 05 11:38:15 <Herrie> JaMa: I just rebooted my machine, I guess that won't work
Jun 05 11:38:23 <Herrie> Seeing otherwise it's stable I suspect a one off issue
Jun 05 11:38:29 <Herrie> If it happens again I'll check dmesg
Jun 05 11:48:14 <JaMa> /var/log/kern.log should still have them
Jun 05 12:27:50 <Herrie> JaMa: Seems no OOM
Jun 05 12:27:52 <Herrie> Weird
Jun 05 12:36:27 <Herrie> Let me just torture the machine once more
**** ENDING LOGGING AT Sat Jun 06 02:59:57 2020