I am having a bunch of weird issues with 6.8.2 that I have never seen before. I am not sure if its just a coincidence and I have cables / drives failing or if there is an issue with my hardware and 6.8.2.

 

Basically I woke up this morning with a disk in my cache pool marked as unmountable with no file system. I rebooted and the issue persisted. I noticed that in the syslog it says a UUID is missing, when I run blkid I can see that the disk now has a different UUID than it did previously.

 

Any idea why this happened? What are the steps for me to restore my cache pool or at least save that data that is on the other disk?

 

tower-diagnostics-20200131-1241.zip

11 minutes ago, johnsanc said:

I rebooted and the issue persisted

Did you by any chance save the diags before rebooting? Those would be the ones showing what happened, also do you know when the pool was created? There was a bug with 6.7.x where new created pools weren't redundant.

Unfortunately I don't but here is the old thread from the last time I had a cache pool issue, which was not long ago. (completely different hardware though)

In the future I will download diagnostics before any reboots. I had syslog saving to the cache pool... but obviously that wont help in this case.

 

You can try these recovery options, but based on current mount error I wouldn't get my hopes up, make sure you try both devices.

Thanks I'll give that a try.

 

I am really struggling with seeing the value of btrfs mirrored cache pool. Seems like every issue that occurs results in a slim chance of recovering data from either drive. There really should be an easy way to at least convert a disk from the pool into a single cache drive. The pool really does give a false sense of security.

Except for the non redundant issue related to the v6.7.x bug, I would say the vast majority of other times data is fully recovered with the options above, but without seeing the syslog from when the problem started can't really guess what happened here, if for example you tried to re-add the dropped device to the pool that would make recovery much more difficult, since Unraid would consider that a new device and wipe it.

Thanks again @johnnie.black - you always seem to come to rescue for things like this. It's a testament to the community and one reason why I stick with unRAID even when weird issues come up every now and then.

 

Thanks to your helpful FAQ I was able to manually mount of the drives, copy data to the array, format the cache pool, and move data back.

 

... Now if only I could figure out why it happened in the first place. If something like this happens again I'll be sure to provide complete diagnostics before rebooting.

Good.

6 hours ago, johnsanc said:

If something like this happens again I'll be sure to provide complete diagnostics before rebooting.

Yep, that's very important.