For a good 3 months now, every day I wake up to my server being offline. The server is still running, so then requires a restart before being able to access it.
I’ve recently tried to rebuild the database and previously tried to systematically switch off scheduled tasks, but that hasn’t helped any and now at around 8:30pm some days the same process needs to be followed.
I have no idea what the logs are trying to tell me, but is there anything there recognised as being the issue?
Your network is very unstable. The messages below appear over 9600 times between ~01:00 and ~07:00.
Verify all the network settings on your server and any associated routers, switches, or other network equipment.
Perform a “network reboot.” Shutdown your server, router, etc. Then, working “from the wall outwards,” power on your router, wait for it to fully come online, proceed with any other network equipment, then finally power on your server.
If that does not help, then try changing the server’s Ethernet cable, using a different router/switch port, etc.
Check if any or your disks with the media is having issues. I had random problems for months, the server became randomly unstable and I had to restart it. And I finally realized a disk had a defective power cable.
Check as well the night jobs configured in your system. Maybe disable all to check if the problem still happens and if the problem is gone bring them back one by one.
I don’t think is the issue, as I’ve also found a large number of these entries in the logs during the times when the server is up and functioning correctly.
To me it feels more like a service failing leaving the server running, but taking access offline.
Is the network configuration of your container and in your host machine complete? Maybe you are missing the gateway or the DNS or both somewhere and that’s why you are getting spammed with those “Network changed” messages.
I got what you were saying, just I’m not so clued up as to Network settings and what not. So DNS and Gateways are a little foreign in my understanding and wouldn’t even know how to check for any issues.
All I know is when the server is running, there are hardly ever any access issues and the host machine is in a different location to where I’m accessing from. Only issues to date are multiple users and bandwidth related.
The network change notices are continuous. The entry DEBUG - Network change. appears over 11,000 times in the ~8 hour span covered by the log files (Plex Media Server.log to Plex Media Server.5.log).
CAVEAT: I’ve never used Unraid, and I’ve never used Docker for Plex.
Plex sees three different interfaces, each on a different subnet (see log entries below).
192.168.0.101
17.18.0.1
17.19.0.1
Network 192.168.0.0 appears to be the local network, as the Internet router is at 192.168.0.1.
There is also traffic to/from 172.17.0.3 (a fourth subnet).
Are all these subnets active on your network?
My only suggestion is to review how the networks are configured for your system.
Hopefully someone more familiar with Unraid and Docker networking can assist.
I’m assuming the 172.X.X.X networks are the Internal Docker Networks where apps on this network can talk directly to each other without going out to the Router. The server I believe has access to these but is not actually running from that network, it’s running from the main Bridge network of the Host, which is direct to the Router where it has the necessary Firewall settings to be accessed Externally. There are 170 mentions of the DEBUG - Network change. in Plex Media Server.log which is the log generated just after restarting this morning and again the server has been working flawless all day.
Where I thought I saw issues could’ve been in Server.2.log, where an issue occurs 66 times before clearing where "threads are waiting on db connections held by threads: 22436653173560, 22436535565112" which then tie back to a Media Metadata Refresh and Database Backup?
Maybe I’ll try the process of elimination with these and the Chapter Thumbnails tasks as a starting point for troubleshooting.
It will verify the integrity of the database, repair if necessary, then re-index it. It is more thorough than the Optimize function in Settings → Troubleshooting.
The readme file has information on running the Utility in a Docker container.
i have a very similar setup (dockerized PMS x2 (Official), running on host network on 2 separate machines (1 is unraid and the other is OMV (Debian Based). i have Zero problems with my server. i just checked my log files and have similar entries ( - at least >1000 / 8H for both the servers ) in my log files.
i suspect your problems are not related to these entries.
huhhhh?!?!