Problems since changing out parity drive and adding SSD to cache pool

[FrefZilla]

I am having trouble getting docker to launch.  Can someone take a look at the attached diagnostics and point me in the right direction?

 

Unraid 6.9.2

tower2-diagnostics-20211127-1034.zip

[trurl]

Cache is full and so docker.img is corrupt. 

 

I think you must have done something wrong trying to add another disk to the pool. One disk is missing from the pool, and the 2 attached SSDs are different sizes.

 

Can you tell us more about what you did?

 

I'll see if @JorgeB has any suggestions.

[FrefZilla]

Cache is remaining full because the log is not clearing during reboot.  After making sure parity was valid after new parity drive, I added smaller SSD to the cache pool.  I thought I had read somewhere that they didn't need to be the same size, but from your response I'm guessing I was wrong.

 

After I encountered errors with the pool, i removed the 2nd SSD.  I ran a balance on the remaining SSD and all seemed fine for a while.  But eventually the docker would become unresponsive.  

[Squid]

Part of your problem is that the cache drive is fully allocated even though it has free space available on it.  A Balance should fix that up.

 

Side note though is that if you're not planning on running a multi-device pool, you're better off using XFS instead of BTRFS for the cache drive as it's more tolerant of abnormal situations.

[trurl]

46 minutes ago, FrefZilla said:

Cache is remaining full because the log is not clearing during reboot. 

Cache doesn't use log space, and log doesn't use cache space. Totally different things. In fact, log space is in RAM.

 

Probably your log is immediately getting full again after reboot due to all the entries caused by cache problems.

[FrefZilla]

I've run balance a few times over the last few days.  Running another one now.

 

How can I change cache to XFS?

 

[trurl]

1 minute ago, FrefZilla said:

How can I change cache to XFS?

reformat

[FrefZilla]

3 minutes ago, trurl said:

Cache doesn't use log space, and log doesn't use cache space. Totally different things. In fact, log space is in RAM.

 

Probably your log is immediately getting full again after reboot due to all the entries caused by cache problems.

Thanks for clarifying.  I think I knew that but have gotten mixed up trying to get back to working.

 

[FrefZilla]

Just now, trurl said:

reformat

I'm sure there is a topic or wiki on it, and I can search for it if you prefer, but what would be the most efficient way to reformat?  Will I lose what is currently on the cache drive (appdata, etc.)?  Any way to keep some/all?  Or would it be most "foolproof" to reformat and start over.

[trurl]

https://wiki.unraid.net/Manual/Storage_Management#Converting_to_a_new_File_System_type