On Wed, Aug 29, 2018 at 11:29:38PM +0200, Niklas Merz wrote:
> Ok my VM with 2 GB RAM for about 20GB data in perkeep should be fine. I
ran
> some uploads and watched the usage and it looked good.

A better way to check if it's getting OOM-killed is via checking kmsg
logs.

The output of "dmesg" will have it if it's recent, and if it's not
recent and has fallen out of the kernel's ring buffer, it should be in
whatever's collecting those logs (`journalctl -k` on many modern
machines).

If it's crashing in certain ways (e.g. segfault), it could also core
dump, at which point it's possible to find a record of that in some
cases. You can find more about this in "man 5 core".

On many modern machines, you can use `coredumpctl` to view any core
dumps that might have happened. To see if that output is accurate, you
should check "/proc/sys/kernel/core_pattern" to see if it references
systemd-coredump, or any other similar tool.


One other thing that might help is the exit code. You said you're
running it via systemd, so the following command should show the exit
code:

$ journalctl UNIT=perkeep.service -t systemd

This should show a line with something like "Main process exited, status=#"
Knowing that number could help a little bit.

You might need to configure your machine further to store kmsg output
and coredumps if it's not configured to, but such changes could help
track down the issue.

- Euan

-- 
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