Ah I see, that's subtle. thanks.

On Sat, 1 Sep 2018 at 19:58, Brad Fitzpatrick <[email protected]> wrote:
>
> If the Go runtime can't get memory from the kernel, Go can crash itself with 
> a log message.
>
> If Go can allocate but the kernel overcommitted itself (on by default) then 
> the kennel picks a process to kill.
>
>
> On Sat, Sep 1, 2018 at 7:52 PM Mathieu Lonjaret <[email protected]> 
> wrote:
>>
>> I make it allocate forever in a loop.
>> Hold on, I'll reproduce it again in a minute.
>> On Sat, 1 Sep 2018 at 19:50, Brad Fitzpatrick <[email protected]> wrote:
>> >
>> > What type of "OOM" are you referring to and how did you cause it? What do 
>> > your logs show when you do so?
>> >
>> > On Sat, Sep 1, 2018 at 7:48 PM Mathieu Lonjaret 
>> > <[email protected]> wrote:
>> >>
>> >> Hmm, so what is it that makes it behave that way? If I force Perkeep
>> >> to OOM (not on systemd), it clearly shows up in the output here. Is my
>> >> kernel not behaving the same way because Perkeep is not running
>> >> through systemd?
>> >>
>> >>
>> >> On Sat, 1 Sep 2018 at 19:44, Brad Fitzpatrick <[email protected]> wrote:
>> >> >
>> >> > The kernel is sending it an uncatchable SIGKILL. It has no chance to 
>> >> > log.
>> >> >
>> >> > On Sat, Sep 1, 2018, 7:18 PM Mathieu Lonjaret 
>> >> > <[email protected]> wrote:
>> >> >>
>> >> >> And you're maintaining that in 'journalctl -u perkeep', at the time of
>> >> >> perkeep's death, there's no such sign of the OOM? I'm asking, because
>> >> >> it usually is pretty obvious just by looking at perkeep's output.
>> >> >> If that's the case, it's something to keep  in mind for future
>> >> >> debugging of Perkeep under systemd.
>> >> >>
>> >> >> On Sat, 1 Sep 2018 at 10:09, Niklas Merz <[email protected]> wrote:
>> >> >> >
>> >> >> > So it failed again.
>> >> >> >
>> >> >> > The exit code is:
>> >> >> > Aug 28 20:21:16 stuff systemd[1]: Started Perkeep Server.
>> >> >> > -- Subject: Unit perkeep.service has finished start-up
>> >> >> > -- Defined-By: systemd
>> >> >> > -- Support: https://www.debian.org/support
>> >> >> > --
>> >> >> > -- Unit perkeep.service has finished starting up.
>> >> >> > --
>> >> >> > -- The start-up result is done.
>> >> >> > Aug 30 16:49:23 stuff systemd[1]: perkeep.service: Main process 
>> >> >> > exited, code=killed, status=9/KILL
>> >> >> > Aug 30 16:49:23 stuff systemd[1]: perkeep.service: Unit entered 
>> >> >> > failed state.
>> >> >> > Aug 30 16:49:23 stuff systemd[1]: perkeep.service: Failed with 
>> >> >> > result 'signal'.
>> >> >> >
>> >> >> > $ dmesg | grep perkeepd
>> >> >> >
>> >> >> > [9339544.204149] perkeepd invoked oom-killer: 
>> >> >> > gfp_mask=0x14000c0(GFP_KERNEL), nodemask=(null),  order=0, 
>> >> >> > oom_score_adj=0
>> >> >> > [9339544.204153] perkeepd cpuset=ns mems_allowed=0
>> >> >> > [9339544.204156] CPU: 3 PID: 11500 Comm: perkeepd Tainted: P         
>> >> >> >   O    4.13.16-2-pve #1
>> >> >> > [9339544.204447] [11488]  1000 11488   459372   282365     638       
>> >> >> > 6        0             0 perkeepd
>> >> >> > [9339544.204450] Memory cgroup out of memory: Kill process 11488 
>> >> >> > (perkeepd) score 539 or sacrifice child
>> >> >> > [9339544.205419] Killed process 11488 (perkeepd) total-vm:1837488kB, 
>> >> >> > anon-rss:1129460kB, file-rss:0kB, shmem-rss:0kB
>> >> >> > [9339544.258384] oom_reaper: reaped process 11488 (perkeepd), now 
>> >> >> > anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
>> >> >> > [9340611.483660] [15711]  1000 15711   682515   276764     707       
>> >> >> > 7        0             0 perkeepd
>> >> >> > [9340611.483664] Memory cgroup out of memory: Kill process 15711 
>> >> >> > (perkeepd) score 529 or sacrifice child
>> >> >> > [9340611.484713] Killed process 15711 (perkeepd) total-vm:2730060kB, 
>> >> >> > anon-rss:1107056kB, file-rss:0kB, shmem-rss:0kB
>> >> >> > [9341096.168476] perkeepd invoked oom-killer: 
>> >> >> > gfp_mask=0x14000c0(GFP_KERNEL), nodemask=(null),  order=0, 
>> >> >> > oom_score_adj=0
>> >> >> > [9341096.168478] perkeepd cpuset=ns mems_allowed=0
>> >> >> > [9341096.168482] CPU: 3 PID: 18250 Comm: perkeepd Tainted: P         
>> >> >> >   O    4.13.16-2-pve #1
>> >> >> > [9341096.168836] [18246]  1000 18246   421195   280089     619       
>> >> >> > 6        0             0 perkeepd
>> >> >> > [9341096.168842] Memory cgroup out of memory: Kill process 18246 
>> >> >> > (perkeepd) score 535 or sacrifice child
>> >> >> > [9341096.169788] Killed process 18246 (perkeepd) total-vm:1684780kB, 
>> >> >> > anon-rss:1120356kB, file-rss:0kB, shmem-rss:0kB
>> >> >> > [9341096.228024] oom_reaper: reaped process 18246 (perkeepd), now 
>> >> >> > anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
>> >> >> > [9501543.900541] perkeepd invoked oom-killer: 
>> >> >> > gfp_mask=0x14000c0(GFP_KERNEL), nodemask=(null),  order=0, 
>> >> >> > oom_score_adj=0
>> >> >> > [9501543.900543] perkeepd cpuset=ns mems_allowed=0
>> >> >> > [9501543.900547] CPU: 3 PID: 22466 Comm: perkeepd Tainted: P         
>> >> >> >   O    4.13.16-2-pve #1
>> >> >> > [9501543.900832] [22465]  1000 22465   412239   265550     648       
>> >> >> > 6    11826             0 perkeepd
>> >> >> > [9501543.900845] Memory cgroup out of memory: Kill process 22465 
>> >> >> > (perkeepd) score 530 or sacrifice child
>> >> >> > [9501543.901793] Killed process 22465 (perkeepd) total-vm:1648956kB, 
>> >> >> > anon-rss:1062200kB, file-rss:0kB, shmem-rss:0kB
>> >> >> > [9501543.956730] oom_reaper: reaped process 22465 (perkeepd), now 
>> >> >> > anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
>> >> >> >
>> >> >> > So it runs out of memory right? I will allocate more RAM to this VM 
>> >> >> > and try running it again.
>> >> >> >
>> >> >> > --
>> >> >> > You received this message because you are subscribed to the Google 
>> >> >> > Groups "Perkeep" group.
>> >> >> > To unsubscribe from this group and stop receiving emails from it, 
>> >> >> > send an email to [email protected].
>> >> >> > For more options, visit https://groups.google.com/d/optout.
>> >> >>
>> >> >> --
>> >> >> You received this message because you are subscribed to the Google 
>> >> >> Groups "Perkeep" group.
>> >> >> To unsubscribe from this group and stop receiving emails from it, send 
>> >> >> an email to [email protected].
>> >> >> For more options, visit https://groups.google.com/d/optout.
>> >> >
>> >> > --
>> >> > You received this message because you are subscribed to the Google 
>> >> > Groups "Perkeep" group.
>> >> > To unsubscribe from this group and stop receiving emails from it, send 
>> >> > an email to [email protected].
>> >> > For more options, visit https://groups.google.com/d/optout.
>> >>
>> >> --
>> >> You received this message because you are subscribed to the Google Groups 
>> >> "Perkeep" group.
>> >> To unsubscribe from this group and stop receiving emails from it, send an 
>> >> email to [email protected].
>> >> For more options, visit https://groups.google.com/d/optout.
>> >
>> > --
>> > You received this message because you are subscribed to the Google Groups 
>> > "Perkeep" group.
>> > To unsubscribe from this group and stop receiving emails from it, send an 
>> > email to [email protected].
>> > For more options, visit https://groups.google.com/d/optout.
>>
>> --
>> You received this message because you are subscribed to the Google Groups 
>> "Perkeep" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to [email protected].
>> For more options, visit https://groups.google.com/d/optout.
>
> --
> You received this message because you are subscribed to the Google Groups 
> "Perkeep" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to [email protected].
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Perkeep" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to