Server fails every few hours on server-qnap

Server Version#: 1.43.0.10346
Player Version#:1.110.0.351-4e48eb83

QNAP TS873a running QTS 5.2.7.3297/64Gb memory

Hardware Transcoder TU117 with 4Gbmemory

Plex Media Server Logs_2025-11-26_07-10-07.zip (742.2 KB)

Plex Media Server Logs_2025-11-27_07-15-28.zip (504.9 KB)

Firstly, apologies, I am new to the forum but not to plex. I have a lifetime plex pass. A very big apology if this isn’t the correct place to report the issue/bug.
I have attached logs and now enabled debug logging, I expect the server to fail in the next few hours. Restarting the package on the qnap app manager allows it to function again, it stops responding on the qnaps address:32400, the web page does not load and no recordings are made.

Recording has operated as exepected until the last two versions of the application package and would appreciate it if anybody could take a look at the log files to point me in the direction of what is failing.

Appreciate the help in advance, especially with the festive seasion around the corner.

Plex Media Server Logs_2025-11-27_14-57-18.zip (4.0 MB)

Logs after 8 eight hours when failure was noticed.

Seeing DNS issues Nov 27, 2025 14:56:39.681 [140106622913336] WARN - [HttpClient/HCl#7] HTTP error requesting GET https://plex.tv/api/v2/user?includeSubscriptions=1&includeProviders=1 (6, Couldn't resolve host name) (Could not resolve host: plex.tv) Still looking at logs

Edit: Server was given a graceful shutdown command Nov 27, 2025 14:55:19.150 [139681674042168] DEBUG - Shutting down with signal 15 (Terminated) Does your QNAP have any kind of resource management?

The graceful shutdown was me restarting the package after I noticed it had stopped recording again. The QNAP has no power management as such except when the UPS tells it to shutdown. nslookup shows that plex.tv can be resolved by the NAS itself at the SSH prompt. 192.168.0.23 is the DNS server, the same result is given if I set the source as the main router.

Interesting, I am seeing multiple Couldn't resolve host name messages. I am also not seeing the normal plex initialization/startup sequence in the logs. Can you stop plex, restart it, wait 10 minutes and post a fresh set of logs?

@a_plex127

In Network & Virtual Switch,

  1. Look at where your primary and secondary DNS points
  2. Temporarily (test), manually set them to 1.1.1.1 and 8.8.8.8\
  3. Now restart PMS

If PMS now works then you know the issue is with the modem/router and/or ISP DNS.

DNS to 1.1.1.1 is fine

There is an error to 8.8.8.8 from even the router, but that isn’t unexpected as it has happened for over a year but 8.8.4.4 does work. PMS failed in a slightly different way, it started recording but was still recording the same program this morning!

Unfortunately I have rolled back the version before I saw your message but plex still had issues with recording (see next message about recording a 30 minute program still running overnight). I have stopped PMS and waited 10 minutes as requested, it has been restarted and the logs have been attached.

Plex Media Server Logs_2025-11-28_11-08-41.zip (3.9 MB)

Any chance you are editing your logs?

If I were then I would have some hair left, no I am not editing the logs. This is a straight upload of the zip files generated by PMS.

My eldest son seems to have corrected the issue, unfortunately I don’t know how but it has involved the cli and the sqlite program running a repair against the database.

He says if it happens again I should just restore the database from a backup… I would do if I knew where it got backed up to….

So, appreciate all the help to this point but for now I will stay cross fingered and hope it stays working over Christmas.

@a_plex127

Did you use my DBRepair tool?

I will ask him next time I am in contact, he doesn’t visit too often, but then again I think it still has the issue but I need to verify the missing program today. I will update as soon as I know more.

Apologies for the delay, my son followed these instructions: https://support.plex.tv/articles/repair-a-corrupted-database/ He may have been faster with your script though!

SInce then I have had one hiccup but that appears to be a vendor EPG issue rather than PMS.

He believes that the database had got too big and got corrupted, I have to go with what he says as I am at the limit of my knowledge.

I believe this can be resolved now unless there is any other input someone wishes to add.