I have been using unraid for a couple years now. But I've never been able to get it completely stable.

I have 2 major problems.

The first being with the prowlarr container. For some reason it always disconnects every few days and I have to manually restart the container and test all indexers to get everything working again for the next few days until it happens again. I think it may have something to do with openvpn that i am running it behind - My assumption is the vpn loses connectivity and then can not connect again after that - saying that, my assumption isn't based on much. I'm not even sure the syslog will show anything regarding this. But maybe someone has some ideas?

 

The second and more major is the constant crashes, which result in me having to create a new USB flash drive and drag/drop the config folder over to get it going again. My server just had the longest runtime of its life of about 40 days, but this time I have had the syslog enabled and backing up to my flash drive so it was recoverable after the crash. However I need help decoding what's actually going on.

I have attached the syslog file, hopefully someone could skim over it and identify the problem(s) I am experiencing.

 

Any help would be greatly appreciated.

syslog

Btrfs is detecting data corruption and there are also some segfaults, I would start by running memtest.

 

Pool should be scrub once this is resolved.

Cheers, appreciate the reply.

Its just running a parity check after the reboot. Once its completed, I'll run a memtest.

Not entirely sure how to perform a scrub, but I'll work it out.

 

I'll post results once completed.

Thanks again

17 minutes ago, Gibbo2787 said:

Not entirely sure how to perform a scrub, but I'll work it out.

Click on the pool device on main, then scroll down to scrub, if it finds errors check the syslog for the list of corrupt files, these should then be deleted/restored from a backup, but only after the issue is fixed, or with bad RAM for example it can give false positives.

I finally got the chance to shut down my server and perform a memtest. Everything looked ok after several passes.

I then performed the scrub on the pool. (Its just the cache right?)

 

This resulted in multiple BTRFS errors and one the says "xhci_hcd 0000:01:00.2: xHC error in resume, USBSTS 0x401, Reinit".

Please see the attached log for the errors. (At the bottom)

 

Can i correct these by enabling "repair corrupted blocks" and performing the scrub again?

 

EDIT* - Tried the scrub with repair corrupted blocks and they appear to be uncorrectable

 

Or should I manually find the file in question and delete it?

 

EDIT* - I just went through and deleted all the corrupted files and its flagging no errors during the scrub now.

I'll see how it goes in the coming weeks/months - hopefully that has sorted out the issue.

Now to sort out the Prowlarr connectivity issue and I won't have to spend so much time checking the server.

 

syslog.txt

Edited March 14, 20242 yr by Gibbo2787

3 hours ago, Gibbo2787 said:

EDIT* - I just went through and deleted all the corrupted files and its flagging no errors during the scrub now.

I'll see how it goes in the coming weeks/months - hopefully that has sorted out the issue.

👍