You are here
Pulsar07 - Tue, 2020/12/15 - 15:15
running the MediaServer V 16.0 as a virutel container in Proxmox.
It eats up all the provided CPU and memory resources, although it is running for several days.
I read that disabling the feature
"Allow remote connections to this Jellyfin Server.", will help, but it does not.
Here is the top output:
top - 14:10:32 up 1:55, 2 users, load average: 0.37, 0.18, 0.14 Tasks: 37 total, 1 running, 36 sleeping, 0 stopped, 0 zombie %Cpu(s): 49.8 us, 0.0 sy, 0.0 ni, 50.2 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st MiB Mem : 4096.0 total, 0.1 free, 4011.3 used, 84.6 buff/cache MiB Swap: 4096.0 total, 3805.8 free, 290.2 used. 84.7 avail Mem PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 187 jellyfin 20 0 7629372 3.8g 22760 S 100.0 96.2 12:55.06 jellyfin
Does anybody have a suggestion, what can help?
Regards
Rainer
Forum:
Hi there, sorry for slow response...
Hi there. Apologies for my slow response...
So if I understand correctly, you are running our v16.0 "Mediaserver" appliance (which includes Jellyfin). You note that you are running it as a container on Proxmox, so I assume a unprivileged LXC container on Proxmox v6.x?
However, you don't note whether this is a fresh install that you are just setting up. Or whether it has been running ok for a while but has recently started this behaviour. The reason why I note that is that if it is scanning a large directory with lots of media in it, then your experience may not be completely unreasonable. On the other hand, if this is an instance that has been running for a while, and this behaviour has just started for no apparent reason, then that suggests an issue...
So is there anything you have done that may have had an impact on the behaviour of your server?
It's also well worth checking both the logs and/or the journal. They perhaps may explain what Jellyfin is doing (or at least trying to do)? IIRC the default log should be found in /var/log/jellyfin. If that doesn't show anything much of value, then you should be able to tweak the log output so that it's more verbose (perhaps enable "debug" logging?). Config files should be in /etc/jellyfin and a quick google suggest that the relevant file to edit is logging.json (careful as it needs to be valid JSON for it to work). Alternatively, you can view the journal like this:
If you can't see any useful, then I suggest that you open 2 separate SSH sessions and watch the journal in one (use the -f|--follow switch to what it live) and restart jellyfin in the other. It might also pay to clear the Jellyfin log so you know that all the contents is related to the fresh start. I.e. in one terminal:
And in the other:
More details ...
installation is a new installation.
it is running as a non privileged LXC container on Proxmox 6.3-3
there exist ~27000 pictures, ~4000 music files, 10 video files
The Mediaserver has 2 i7-8559U CPU @ 2.70GHz assigned and 4GB memory.
The server was running for about 5 days. The memory is completly used after some minutes and one CPU is at 100% all the time.
In the journal there are some single exception at startup, because of file permissions and file format issues, but at the end there is endless reappearing message with no hint for me:
That error looks like something up with XML?
Thanks for the details. I'll launch one myself and have a look to see if I can reproduce on my end.
That error every 10 seconds doesn't look ideal, but I wouldn't have expected that to cause it to max out a CPU core and run out of RAM?! My guess is that there's something else going on.
Regardless of whether it's the cause of your issues or not, I did a quick bit of googling re your errors and came across a post that mentioned a case of DLNA device that was giving malformed XML. I can't find the link now though and I have no idea how you'd go about testing/diagnosing that possibility...
Out of interest, has your root volume got plenty of free space?
Add new comment