[SOLVED] Cache Drives dropped off System

[Cessquill]

Hi - just powered down my system to make space for a new drive.  When I powered back on, one of the cache drives had become unplugged.

 

Powered back down, reseated the power cable to the cache drive and powered back up again.

 

Now I have no cache, but my two cache drives are in the unassigned devices section.

 

I had two 250 GB drives pooled to give 500 GB of cache in a BTRFS.

 

Is it possible to remount them as cache and not lose data, or will I lose it all?  (I can recreate the docker image, and have appdata backed up)

 

What's the best way to proceed?  I'm running 6.8.3

[JorgeB]

Did you start the array with a single cache device assigned?

[Cessquill]

12 minutes ago, johnnie.black said:

Did you start the array with a single cache device assigned?

When one of them was unhooked, yes.

 

Powered back down, plugged in, turned on, neither assigned.

Edited July 28, 2020 by Cessquill

[JorgeB]

2 minutes ago, Cessquill said:

When one of them was unhooked, yes.

That might be a problem, do you have those diags?

[JorgeB]

If not, do you remember if the cache mounted? Is the pool raid1?

[Cessquill]

2 minutes ago, johnnie.black said:

If not, do you remember if the cache mounted? Is the pool raid1?

I don't have diagnostics - didn't think to grab them at the time I'm afraid.  The cache didn't mount as far as I am aware, as the docker service couldn't start.

 

I do have my disk assignments file (from appdata backup?) so I at least know which was cache and which was cache 2.

 

I only built the cache once when I set it up, and am struggling to remember what I did (might have been covered in a Space Invader video).  What I know though - it was set up as 240 GB and 250 GB  drives pooled to give me 490 GB of cache.  Does that help?

[JorgeB]

4 minutes ago, Cessquill said:

Does that help?

Yes, that and that it didn't mount, it should be OK then, do this:

 

Stop the array, if Docker/VM services are using the cache pool disable them, unassign all cache devices, start array to make Unraid "forget" current cache config, stop array, reassign all cache devices (there can't be an "All existing data on this device will be OVERWRITTEN when array is Started" warning for any cache device), re-enable Docker/VMs if needed, start array.

 

P.S. stopping docker/VMs won't matter since array was already started without a pool, but since I copy/paste the reply it's complete.

[Cessquill]

Looking at it, my last appdata backup ran at 3:00 this morning, and I can recreate the img file in no time.  The only other thing on the cache were files waiting to be written to the array (which is why I have a new drive on order) - I can get to that and pull them off by mounting the drives.

 

It shouldn't matter then if the cache is wiped, hopefully.  Just need to recreate the same pool.

[JorgeB]

If you do the above it should mount correctly.

[Cessquill]

3 minutes ago, johnnie.black said:

If you do the above it should mount correctly.

Thanks - our posts crossed.  I'll try that.

[Cessquill]

32 minutes ago, johnnie.black said:

If you do the above it should mount correctly.

Perfect!  THANK YOU SO MUCH!

 

Worked first time, and everything is back to how it was.  I'm running in single mode at the moment, but will likely switch to two pools once 6.9 comes out.  Eventually maybe 3 and give Plex its own appdata.

 

All of my array drives are in a hotswap area, but the curse of loose power cables is still strong for utility drives inside the case...

 

Thank you again.