Hi,

 

I had to hard reset my server as it was hanging and now my cache pool seems to be corrupt - it goes into read only mode

 

Here is the disk info after a check:

Quote

 

[1/7] checking root items [2/7] checking extents corrupt extent record: key [118557515776,168,20480] bad key ordering 14 15 bad key ordering 14 15 bad key ordering 14 15 bad block 287457280 ERROR: errors found in extent allocation tree or chunk allocation [3/7] checking free space cache [4/7] checking fs roots [5/7] checking only csums items (without verifying data) there are no extents for csum range 1879971414016-1880016683008 csum exists for 1879970152448-1880016683008 but there is no extent record ERROR: errors found in csum tree [6/7] checking root refs [7/7] checking quota groups skipped (not enabled on this FS) Opening filesystem to check... Checking filesystem on /dev/sdb1 UUID: 9d23ae79-f5d0-4b39-8941-065914c61534 found 77987770368 bytes used, error(s) found total csum bytes: 1375216 total tree bytes: 280264704 total fs tree bytes: 238010368 total extent tree bytes: 39157760 btree space waste bytes: 80888550 file data blocks allocated: 259743842304 referenced 66393468928

 

 

I've attached my server diagnostics as well. Do I need to format the cache pool? If so, how do I do this with a cache pool? I've copied off everything I need from the pool and have backups from last night of appdata, system shares

server-unraid-diagnostics-20200414-1601.zip

17 minutes ago, ultimz said:

Do I need to format the cache pool?

It's the best bet, backup, format and restore the data, there some recovery options here if needed.

Sorry just read the readme for this forum and realise that I'm missing some info

 

Unraid Version: 6.8.3 with Nvidia drivers

I've got 2 500GB SSD drives in the cache pool and 5 SATA drives in the normal array. I am running an ASROCK motherboard with 4 ports and am using the external SATA connector for a 5th port. I then have a mini pci-e card for 2 more SATA ports

 

All I've done so far is try to run the mover to copy data off the cache without success (along with a few reboots to get write working again - it works but then stops when I think it hits the Plex appdata folder... not sure). I have then manually copied the data from the cache to a share on the array (hope it's still fine) and then run the diagnostics tests

1 minute ago, johnnie.black said:

It's the best bet, backup, format and restore the data, there some recovery options here if needed.

Thanks Johnnie - I am trying to do this but for the life of me can't figure it out and don't want to break things further. Will this work:

 

Stop the array

Remove both cache disks from the cache pool (from the "Main" screen)

Use unassigned devices plugin to delete the partitions on both cache disks

Add them back in and hopefully format?

Assuming backups are done and trim is supported by the controller used the easiest way is with the array stopped type on the console:

 

blkdiscard /dev/sdX

 

Do that for both, restart array and format.

Ok I have moved both SSDs to my main motherboard controller (not the PCI-E card) and have run that command and formatted the cache pool. Seems like it's fine and busy copying data back on to it. Will try and start up the docker service after this... hope I didn't break anything (wow this is stressful)

 

Thanks for all the help @johnnie.black. I will let you know if I run into any issues

Looks like my appdata folder is corrupt so restoring from backup and busy testing docker containers. Not a big deal! Everything else looks good

 

Thanks @johnnie.black I really appreciate all the help on this