Plex server stops responding (both web interface and server) several times a day (+docker)

General Discussions
1

I’m running the plexpass version of plex on docker on an Ubuntu 16.04.3 LTS host. The /data and /config directories are mounted over NFS from a NAS.

After a container restart, it starts off just fine. A little while later, the health-check reports that the server is unhealthy. And, accessing the web interface on port 32400 doesn’t work. Neither do any plex clients.

I turned on verbose server debugging, but couldn’t see anything obvious. According to log, it did seem that the Plex Relay was still running. It also seemed that one of my Roku’s had a paused movie that seems to be pinging the server (somehow successfully?)

Anyways, all the usual Plex Processes seem to be running, its just not servicing any new clients or the web interface.

As an aside, another bug with the docker health-check script is that the --connect-timeout 15 isn’t working. That is because the connection succeeds. There are just no responses. Instead of --connect-timeout, --max-time should be used instead.

2

See this recent response to lockouts when using nfs

3

Ah, interesting. I’m going to try this out.

In case it works, I think the Plex docker README.MD should be updated to encourage users to NOT use bind mounts/NFS for the /config directory, but instead, use container volumes.

In bold red letters maybe.

4

Oh man, I should RTFM - it is actually in the README.MD. Could be a little more prominent tho :wink: