September 22, 20196 yr Hi all, Hoping somebody can assist here, I think all data on my cache drives has been lost. I was having issues with UnRAID popping up errors claiming the file system was read-only so I restarted. Following this, the terminal itself came up with an error: 'BTRFS critical, corrupt leaf, bad key order' and lots of numbers. When trying to start the array it claimed that one of the dual SSD caches was unmountable. There was a little format button next to the drive (not the cache array itself) which I clicked, assuming that the cache drive would be rebuilt from the second drive. However it appears that this has cleaned everything out, there is now nothing in /mnt/cache, and all of my dockers and VMs are gone. Oops. Do you know how I can rebuild the cache using the second drive? I've been using Duplicati to backup but stupidly I hadn't checked it was working and it has been stuck for over a month. Any advice would be appreciated as I'm a bit stuffed now - everything pretty much was on those cache drives and if I can't get to the data I'm really stuffed. Cheers
September 22, 20196 yr Community Expert Go to Tools-diagnostics and attach the complete Diagnostics zip file to your next post
September 22, 20196 yr Author No problem - please see attached. I've mounted the cache 2 SSD using cryptsetup and I'm currently coping all the files from there to another SSD in the server (it seems to be missing tons for some reason?) and the SSD that the original issue I've removed and am currently doing a drive dump on so there isn't SMART logs on there for that one. It was all fine prior to the issue though, no errors. Please let me know if you need anything else, happy to plug back in once the dumps are done if you need. tower-diagnostics-20190922-1307.zip Edited September 22, 20196 yr by ThinkingGiraffe
September 22, 20196 yr Community Expert From the FAQ: https://forums.unraid.net/topic/46802-faq-for-unraid-v6/page/2/#comment-543490 Lots of other things about managing the cache pool in the FAQ.
Archived
This topic is now archived and is closed to further replies.