[SOLVED] Unable to remove old cache drive

[Chuck Claunch]

I used to run with two cache drives, a Samsung 830 and a PNY, both SSDs.  I decided I didn't need both as cache devices anymore and removed the PNY drive as cache.  It now shows up in unassigned devices (along with a busted jpg icon).  If I physically remove the PNY drive from my system, my remaining Samsung cache drive says unreadable.  If I physically put the PNY drive back in, the Samsung cache drive is readable again and the PNY drive still shows up in unassigned devices with the busted icon.  See screenshot.  How can I release this drive so I can remove it and not lose my cache contents?  All of my docker and VM configs are on this and they're totally missing if I boot without the PNY drive (and Samsung cache drive says unreadable).

[trurl]

 

[JorgeB]

Before doing anything else please post the diagnostics: Tools -> Diagnostics

[Chuck Claunch]

Here ya go

unraid-diagnostics-20210825-0707.zip

[JorgeB]

Pool is using single profile for metadata so it failed to remove the device, likely it was created with v6.7.x, there was a bug, do this to fix:

 

Stop the array, if Docker/VM services are using the cache pool disable them, unassign the single cache device, start array to make Unraid "forget" current cache config, stop array, reassign both 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, convert pool to raid1 and once that's done remove cache2 again.

 

 

[Chuck Claunch]

Ok so just so I'm clear, if I disable Docker/VM services I won't lose all that data?  Will it just go over to the main pool?

[JorgeB]

9 minutes ago, Chuck Claunch said:

if I disable Docker/VM services I won't lose all that data?

You won't lose the data, but they won't go to the main pool, it's just for when you start the array without cache Unraid doesn't create new docker/libvirt images, also make sure cache backups are up to date before doing this, not a risky operation but just in case.

[Chuck Claunch]

Ok thanks! will give this a try this evening

[Chuck Claunch]

Thanks a ton @JorgeB the steps you outlined worked perfectly 👍

[vvolfpack]

I'm in the same boat as OP wherein my server doesn't boot if I remove the NVME drive (which is currently unassigned) from the slot. It's not in a cache pool with the primary m.2 device. Before following the same steps by JorgeB to troubleshoot, I wanted to check with the forum if I can proceed. FWIW, I'm going through a full migration of the server to a new MoBo, Intel CPU, and a new cache drive.

vvolfbox-diagnostics-20221105-1812.zip

[JorgeB]

10 hours ago, vvolfpack said:

I'm in the same boat as OP wherein my server doesn't boot if I remove the NVME drive (which is currently unassigned) from the slot.

Do you mean the pool does not mount as was the problem for the OP or the server does not boot? I guess the latter since the unassigned NVMe device is not using the same filesystem.

[vvolfpack]

It's the latter! Server does not boot once I removed the unassigned NVMe device from its slot

[ConnerVT]

Syslog Server running, saving file to NVMe?

 

(Just spitballing here)

[ChatNoir]

When is the server failing to boot ?

During Unraid's boot process or before ?

[vvolfpack]

 

On 11/6/2022 at 1:58 PM, ConnerVT said:

Syslog Server running, saving file to NVMe

Thank you for all your comments! Turns out that Syslog was indeed running on the NVMe. I was able to resolve it. Thanks!