So I went to use plex today and realized it couldn't connect to my server. I looked at the server and everything appeared fine. After some more searching I found this in the logs.

 

ErrorWarningSystemArrayLogin

Feb 14 17:10:19 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 766203 off 0 csum 0x6330a5c2 expected csum 0x00000000 mirror 1
Feb 14 17:10:19 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:19 Tower kernel: BTRFS info (device sdf1): no csum found for inode 766203 start 0
Feb 14 17:10:19 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 766203 off 0 csum 0x6330a5c2 expected csum 0x00000000 mirror 2
Feb 14 17:10:19 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:19 Tower kernel: BTRFS info (device sdf1): no csum found for inode 766203 start 0
Feb 14 17:10:19 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 766203 off 0 csum 0x6330a5c2 expected csum 0x00000000 mirror 1
Feb 14 17:10:19 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:19 Tower kernel: BTRFS info (device sdf1): no csum found for inode 766203 start 0
Feb 14 17:10:19 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 766203 off 0 csum 0x6330a5c2 expected csum 0x00000000 mirror 2
Feb 14 17:10:19 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:19 Tower kernel: BTRFS info (device sdf1): no csum found for inode 765591 start 0
Feb 14 17:10:19 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 765591 off 0 csum 0xea2f52b5 expected csum 0x00000000 mirror 1
Feb 14 17:10:19 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:19 Tower kernel: BTRFS info (device sdf1): no csum found for inode 765591 start 0
Feb 14 17:10:19 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 765591 off 0 csum 0xea2f52b5 expected csum 0x00000000 mirror 2
Feb 14 17:10:20 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:20 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:24 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:24 Tower kernel: BTRFS info (device sdf1): no csum found for inode 766203 start 0
Feb 14 17:10:24 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 766203 off 0 csum 0x6330a5c2 expected csum 0x00000000 mirror 1
Feb 14 17:10:24 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:24 Tower kernel: BTRFS info (device sdf1): no csum found for inode 766203 start 0
Feb 14 17:10:24 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 766203 off 0 csum 0x6330a5c2 expected csum 0x00000000 mirror 2
Feb 14 17:10:24 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:24 Tower kernel: BTRFS info (device sdf1): no csum found for inode 766203 start 0
Feb 14 17:10:24 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 766203 off 0 csum 0x6330a5c2 expected csum 0x00000000 mirror 1
Feb 14 17:10:24 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:24 Tower kernel: BTRFS info (device sdf1): no csum found for inode 766203 start 0
Feb 14 17:10:24 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 766203 off 0 csum 0x6330a5c2 expected csum 0x00000000 mirror 2
Feb 14 17:10:24 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:24 Tower kernel: BTRFS info (device sdf1): no csum found for inode 765591 start 0
Feb 14 17:10:24 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 765591 off 0 csum 0xea2f52b5 expected csum 0x00000000 mirror 2
Feb 14 17:10:24 Tower kernel: BTRFS critical (device sdf1): corrupt node: root=7 block=772802199552 slot=24, bad key order, current (18446744073709551606 128 70054182912) next (18446744073709551606 128 1344667648)
Feb 14 17:10:24 Tower kernel: BTRFS info (device sdf1): no csum found for inode 765591 start 0
Feb 14 17:10:24 Tower kernel: BTRFS warning (device sdf1): csum failed root 5 ino 765591 off 0 csum 0xea2f52b5 expected csum 0x00000000 mirror 1
 

 

 

It just keeps on going with errors. I have 2 ssd's running in my cache and don't know where to go from here. I was able to copy some of the files from my cache to my other computer as a backup, but although I can view them all, I can't copy some of them. The actual array seems fine. I'm not super savy with unraid so please any help is appriciated. My work files (quickbooks) are on this cache (I have current backups on another computer so I'm atleast protected there) so I need it up and running asap.

 

Thanks

 

 

EDIT: I deleted a lot of the repeating error code to make it easier to read. It just keeps throwing those errors. See post below for diagnostic zip

Edited February 15, 20215 yr by Steve Welsh

Sorry for that being so long. Only way I could copy and paste was to copy it all. Here's the diagnostic zip file. Thanks!

tower-diagnostics-20210214-1726.zip

Cache filesystem is severely corrupt, you should backup and re-format, also good idea to run memtest to check the RAM for errors.

I backed up what I could but stupid question, how do I format the cache drive? I deleted everything I could off it (can't seem to delete the plex folder so I guess that's where my issue is). I've looked all over and can't find a format option. 

 

Thanks

You can jus wipe them and them Unraid will offer to format, with the array stopped wipe both with:

 

blkdiscard /dev/sdX

 

Start array and format.

Thanks! Did that and just finished loading backups. I'm back up and running. Running the memtester86 now and getting a lot of errors (464 errors on test 5 and 8 so far). Still running. So guessing my memory is the problem? Anyone recommend a particular memory brand or type (my motherboard isn't compatible with ECC memory)? I have corsair in it now. In all my years of running a windows PC, never had memory fail on me. I'm not very savy with unraid but this memory is only 2 years old so a little surprised.

 

Thanks again for the help!

Edited February 16, 20215 yr by Steve Welsh

43 minutes ago, Steve Welsh said:

Running the memtester86 now and getting a lot of errors (464 errors on test 5 and 8 so far). Still running. So guessing my memory is the problem?

Yep. Zero errors is the only acceptable result. Try testing one stick at a time and see if you can narrow it down.

Thanks. First stick threw the codes. Second stick has been running for an hour so far and no codes yet so looks like I found the problem. Going to let it run overnight and if no codes are thrown I'll probably just leave it run on the one stick (8gig) and see if I need more. Thanks for the help everyone!