So I think the source of many problems I've had lately is my first cache drive (cache 1). I have two 480GB SSD's installed.

 

Today, again, cache 1 disappeared from the cache pool entirely. When I stopped the array, it was showing as unassigned. Now, no matter what I do with cache 2 I can't get the pool to start up again. I've tried unassigning cache 1, unplugging cache 1, changing cache 2 to cache 1 in the config, changing down to a single cache pool, everything I can think of. No matter what, the pool is unmountable.

 

Two questions:

1. How do I get my pool back up and running?

2. What good is two cache drives in a RAID 1 pool if the other can't just step in for the broken one?

 

Thanks.

Since you have more than one cache disk (cache pool), you probably want to replace your failed cache1 disk with a new disk to keep it a pool.  If you want to go to a single cache drive, stop the array and decrease the cache slots to one.

Lots of information about managing cache pools in the V6 FAQ sticky at the top of this subforum.

5 minutes ago, dboonthego said:

Since you have more than one cache disk (cache pool), you probably want to replace your failed cache1 disk with a new disk to keep it a pool.  If you want to go to a single cache drive, stop the array and decrease the cache slots to one.

 

I tried that. It said it was unmountable.

 

Because the good drive is on slot 2 in the pool, you can't decrease the size to 1 without first moving the "good" drive to slot 1.

 

Give this video a look:

 

1 hour ago, ksignorini said:

Because the good drive is on slot 2 in the pool, you can't decrease the size to 1 without first moving the "good" drive to slot 1.

 

Yep, that should be all you need to do, then reduce the slots down to 1. 

Set Cache2 = no device

Set Cache1 = The previous cache2 disk

Set the slots = 1

Start array

 

1 hour ago, ksignorini said:

no matter what I do with cache 2 I can't get the pool to start up again.

Side note.. What version of unRAID are you running?  The system should run normally if a cache pool disk fails as of V6

Edited February 28, 20179 yr by dboonthego

4 minutes ago, dboonthego said:

Side note.. What version of unRAID are you running?  The system should run normally if a cache pool disk fails as of V6

 

6.3.2

 

And, unfortunately, it didn't run properly. And after doing the above, it didn't allow me to mount the cache 2 as new cache 1.

 

I did manage to get my pool up and running, read-only at least, by putting that drive on another port and got much of the essential stuff (appdata) "moved" onto the array. However, it's crapping out again (* for temp, billions of reads reported on that drive) and so I'm sure that SSD it kaput. I plan to rebuild the pool with just the one good cache until I get this one replaced on warranty.

For anyone interested, something was massively wrong with my cache pool, probably caused by the bad drive which is now gone back to the manufacturer for replacement.

 

What I ended up doing was rebuilding my cache drive from scratch. Luckily, I use the Community Applications backup to keep a copy of my appdata on my array. I also had some recent (though not perfect) backups of my VM's which I used to rebuild nearly complete versions on the new cache drive.

 

When my other SSD comes back from RMA, I'll get it back into the machine and into the pool I guess.

 

I'm still wondering, despite what I've read, with a standard cache pool using RAID 1, shouldn't one of the drives be complete stand-in for the other if one goes bad?

16 hours ago, ksignorini said:

I'm still wondering, despite what I've read, with a standard cache pool using RAID 1, shouldn't one of the drives be complete stand-in for the other if one goes bad?

 

Yes.  Not sure what happened in your situation though.

5 hours ago, dboonthego said:

 

Yes.  Not sure what happened in your situation though.

 

Yeah, and that's the rub. Oh well. Rebuilt and all is fine now.