My Plex server crashes every night, All I know at the moment (due to TrueNAS guys claiming it was due to SWAP and I should increase RAM, so increased from 16GB → 32GB now), but still.
I can see there is a spike in Swap usage just after 3am, but can’t say if thats the cause of the crash, or time, I’ve attached my Plex logs…
Diagram attached.
Dec 15 03:13:05 vaultx kernel: pid 2010 (Plex Media Server), jid 1, uid 972, was killed: out of swap space
Dec 15 03:13:05 vaultx kernel[1095]: Last message 'pid 2010 (Plex Media' repeated 1 times, suppressed by syslog-ng on vaultx.xxxxxxx.com
Dec 15 09:25:59 vaultx kernel: pid 32784 (Plex Media Server), jid 5, uid 972: exited on signal 6
Dec 16 03:15:06 vaultx kernel: pid 44663 (Plex Media Server), jid 7, uid 972, was killed: out of swap space
Dec 16 03:15:06 vaultx kernel[1095]: Last message 'pid 44663 (Plex Medi' repeated 1 times, suppressed by syslog-ng on vaultx.xxxxxxx.com
So all the media is on TANK, apps are on Apps.
I\m also attaching a Swap usage diagram, you will see swap usage was always zero, up to a couple of weeks ago when a Plex upgrade was done.
RAM use to be 16GB, and stable, upgraded on Tuesday to 32GB after everyone implied this is a lack of RAM leading to SWAP leading to crash… my issue, my media volume have not grown significantly…
The problem of out of Swap, I think it is a symptom, it’s not the cause.
Do you have a snapshot you can roll back to to verify that the older PMS version worked?
I don’t have access to my TrueNAS install currently (power outage) but is it possible that the jail is running out of swap space, and not TrueNAS itself?
Also, seems like the swap issue happens around 3AM, which I think might be the default time for the daily housekeeping tasks, etc… You could try disabling/re-scheduling this to see if that is having an impact as well.
So I moved the the normal house keeping schedule from 2-5 → 4-7 and likewise my Plex crash/kill moved…
I’m a full Plex Pass pain up member, surely there is some form of official ticket system ?
root@vaultx[~]# cat /var/log/messages |grep Plex
Dec 17 05:14:35 vaultx kernel: pid 2020 (Plex Media Server), jid 1, uid 972, was killed: out of swap space
Dec 17 05:15:34 vaultx kernel[1100]: Last message 'pid 2020 (Plex Media' repeated 1 times, suppressed by syslog-ng on vaultx.xxxx.com
root@vaultx[~]#
You could check if this is the problem that is caused by having loads of extras (trailers/clips) in the database for movies - that issue was leading to memory allocation failures and crashes - there is no evidence of memory allocation failures in the Plex Media Server logs but still worth checking the database - running a query like this on the database
select count(metadata_item_id),metadata_item_id from metadata_relations group by metadata_item_id order by count(metadata_item_id) desc limit 20
Unfortunately we do not have crash reporting and mini dump creation on FreeBSD
But instead of repair - run this sql select count(metadata_item_id),metadata_item_id from metadata_relations group by metadata_item_id order by count(metadata_item_id) desc limit 20
missing something, that directory looks like it will be inside my jail
so… I can’t access it if I’ve shut the rail down, which in my case imply shutting Plex down.
Will keep it running as I’m only going to run a select
OK - it is not that straightforward - the db filename needs to be a parameter to the plex sqlite binary
So you navigate current directory ito /Plex Media Server/Plug-in Support/Databases
then you launch “Plex SQLite” binary giving it this as first parameter com.plexapp.plugins.library.db
and that goes into sqlite command line where you can type select count(metadata_item_id),metadata_item_id from metadata_relations group by metadata_item_id order by count(metadata_item_id) desc limit 20
It’s running on TrueNAS CORE atm (which is BSD based of course)… been thinking about seeing whats involved to move to TrueNAS SCALE (which is linux based)…
bit nervous of the move, but might end being the better option in the absence of any other.
