UnRAID Server Version: 7.0.1

PMS: 1.41.5.9522

Hardware: Beelink S12 Pro

 

Hi there, I have a frustrating problem in that anytime I ask Plex to scan for new media or it's doing routine maintenance checks overnight, it crashes my entire UnRAID server and only started on the weekend - everything worked fine as of Mar 8. Since then, an updated for PMS was released which I hoped would fix things but the issue persisted.

 

Once you click the "scan" button, the entire system becomes unresponsive and I can't access the WebGUI anymore. It also freezes any operations currently running (I was doing a data back up at the time of one of the crashes). If I leave the server running overnight, once the maintenance kicks in, it also freezes everything.

 

Here's what I've tried so far based on other forum posts across UnRAID, Plex and reddit:

• Repaired the database
• Switched to ipvlan from macvlan
• Deleted "codecs" folder for Plex
• Re-installed Plex and tried 2 different database/meta data back-ups
  • 1 back up was an older one when things were working perfectly fine
• Validated all permissions and owners for the appdata on my cache drive
  • Owner is "nobody"
  • Permission is full read/write for every folder

 

Attached are the PMS and UnRAID logs. 

 

I'm fairly new to Linux so all I could really decipher is that something is causing a problem with nginx which I believe is tied to the WebGUI but couldn't see anything regarding Plex.

 

I have posted this issue on the Plex forums as well but haven't received any responses.

 

Any help would be greatly appreciated!

Plex Media Server Logs_2025-03-13_13-13-15.zip syslog-192.168.2.83.log

Attach Diagnostics to your NEXT post in this thread.

3 minutes ago, Jefflix said:

Beelink S12 Pro

USB not recommended for assigned disks for many reasons.

Acknowledged on the USB, but to be fair, it hasn't been a problem until now and my previous setup was on a TerraMaster NAS w/ 2 DAS via powered USB hub which also didn't have any issues.

 

**just editing to add context in that I don't mean to come off as defensive given I'm the newbie here. Based on recos before moving from TOS to UnRAID, I'm not running any sort of parity, all drives are operating independently in a JBOD configuration. Not sure if that change anything.

 

Diagnostics attached.

jefflix-diagnostics-20250313-1336.zip

Edited March 13, 20251 yr by Jefflix
Added context

1 hour ago, Jefflix said:

I'm not running any sort of parity, all drives are operating independently in a JBOD configuration.

That is what I usually recommend. Then nothing can get out-of-sync if something disconnects.

 

Also looks like drives are identified correctly, and you get SMART reports, so you have a good USB configuration. Doesn't always work out that way for some people.

Have you done memtest?

Whew, glad I didn't set myself up for failure!

 

Just completed 1 pass on memtest and it passed with 0 errors. I'm letting it complete 2-3 passes just to be safe though. Will let you know if anything changes.

 

**Edit: Completed 2 passes successfully with 0 errors, took 1h43m to complete

 

 

Edited March 13, 20251 yr by Jefflix

Another update, I tried re-doing the array; did a "New Config" and re-inserted each drive into the exact same order. I was then able to scan each library, 1 by 1 in their entirety. Then I tried the "Scan All Libraries" and then it froze the system again.

 

Also to note, the "sdX" assignments seem to have changed from when I first installed the drives vs. after the reboot following the crash. I only noticed because Drive 9 for some reason needs me to manually enable write cache after every reboot. After re-doing the array it was sdk, now it's sdi. 

sdX designations aren't expected to be the same, and will often change, especially when adding, removing, or replacing disks.

 

Unraid keeps track of disk assignments by serial number.

 

I often recommend not even thinking of sdX and instead use the disk assignments, or specify the last 4 characters of serial when we need to discuss specific disks.

22 minutes ago, Jefflix said:

I was then able to scan each library, 1 by 1 in their entirety. Then I tried the "Scan All Libraries" and then it froze the system again.

I notice you seem to have User Shares named for each disk. That's OK, but the point of user shares is it allows folders to span disks.

 

So do you have plex configured with a separate library for each disk?

Gotcha re: sdX!

 

I setup individual shares to help me keep things organized. Kind of a relic of how I had to do things when using the TerraMaster system but am used to it now. There are only a few Plex libraries though that are comprised of folders on each drive more or less. For example, most drives have a "Movies" folder so the Movies library on Plex os directed to each one.

Post docker run for plex

So you have /mnt/user (or possibly even /mnt) as a host volume for plex container, and then when you create a library in Plex, you add multiple folders to the library.

 

When scanning a single library then, does it actually involve multiple disks?

Here ya go!

 

re: scanning library query; yes, the host volume is /mnt and the scan would span multiple drives. 9 drives for 2 of the libraries, most other libraries are limited to 1 drive. 

Docker Run - Plex.txt

 

Edit: wait... may have posted the wrong log.

Docker Run - Plex.txt is from going to the main UnRAID console and using "docker logs Plex-Media-Server"

Plex Logs.txt is from going to the Docker tab, clicking on the Plex icon and selecting "Logs"

Plex Logs.txt

Edited March 14, 20251 yr by Jefflix

Wasn't really asking for logs.

 

Some words or phrases in the forum are automatically converted to links which go to explanations of the word or phrase, such as docker run and diagnostics.

Oh sorry lol def a noob

 

What did you mean by "post docker run for plex"?

5 minutes ago, trurl said:

Some words or phrases in the forum are automatically converted to links which go to explanations of the word or phrase

Did you click the link?

Apologies, got caught up on the latter part of the post, here's the run command. I put x's for the claim code, everything else is as-is.

run.txt

That confirms what we had already discussed.

38 minutes ago, Jefflix said:

the host volume is /mnt

I just decided to pursue it as a teaching moment.

 

What I was wondering about is whether or not plex tries to access multiple disks (through your single USB connection) when scanning all libraries.

Np, appreciate that!

 

Yes, the mini PC is connected to a powered USB 3.0 hub with 3 DAS units. Before using the Beelink, the NAS had 2 DAS units connected to it with the same powered USB hub.

Lots of call traces in that syslog you posted at the beginning, but no idea what might be the cause. Maybe someone else will chime in.