Replacing cache drive gone wrong

[RockDawg]

I have a Crucial MX500 250GB SSD that I was using as a cache drive and it started reporting a SMART error so I wanted to remove it and investigate further.  I had a spare Samsung EVO 860 that I decided to replace it with.  I've never used the cache for anything other than Docker and VMs.  Never used mover.  So I follow the instruction in the wiki and mover empties my Crucial as expected. 

 

The first issue arose in that after mover put everything back on my new Samasung cache drive and I rebuilt the Docker image Unraid show 211 GB used on the new cache drive when it was only 133 GB before the swap.  Don't know why this happened.

 

Next I couldn't access my Docker containers' pages and then I realized that my custom Docker network was gone.  I created a new one and assigned it to my containers and that was fixed.  Don't know why this happened.

 

I have two other drives I mounted via Unassigned Devices (Temp_Intermediate and Temp_Download).  Now neither will mount.  The log shows the following when I try to mount one:

 

Sep 2 20:18:50 Tower unassigned.devices: Adding disk '/dev/sdx1'...
Sep 2 20:18:50 Tower unassigned.devices: Mount drive command: /sbin/mount -t xfs -o rw,noatime,nodiratime,discard '/dev/sdx1' '/mnt/disks/Temp_Intermediate'
Sep 2 20:18:50 Tower kernel: XFS (sdx1): Filesystem has duplicate UUID 351d1423-92bc-4563-854d-9ac562958721 - can't mount
Sep 2 20:18:50 Tower unassigned.devices: Mount of '/dev/sdx1' failed. Error message: mount: /mnt/disks/Temp_Intermediate: wrong fs type, bad option, bad superblock on /dev/sdx1, missing codepage or helper program, or other error.
Sep 2 20:18:50 Tower unassigned.devices: Partition 'Samsung_SSD_860_EVO_250GB_S3YHNF0M702349N' could not be mounted.
Sep 2 20:18:50 Tower unassigned.devices: Don't spin down device '/dev/sdx'.

 

So neither will mount now, but Temp_Download still shows as a share and I can access it's contents from my Windows laptop. 

 

So what the heck is going on with all these issues and how do I fox this last issue?

So both sho

[RockDawg]

Rebooting unraid fixed the issue with unassigned drives.  But now I have issues on a couple Docker containers.  Radarr is not the V3 I was running and Sonarr will not launch the WebUI.  THis combined with all the extra cache drive space being used now and this has been anything but a smooth and painless transition. 

 

I'm just not sure what I did wrong.

[JorgeB]

6 hours ago, RockDawg said:

Unraid show 211 GB used on the new cache drive when it was only 133 GB before the swap.  Don't know why this happened.

If you have vdisks there mover won't keep them sparse, you can use cp --sparse=always to "resparsify"

 

 

[RockDawg]

So do I copy the vdisk off the cache drive and then copy it back with that switch?

[trurl]

Post diagnostics.

[RockDawg]

Diagnostics attached.

tower-diagnostics-20200903-1847.zip

[JorgeB]

11 hours ago, RockDawg said:

So do I copy the vdisk off the cache drive and then copy it back with that switch?

Probably the best, you could also do it in place then delete the old one but might not have enough space.

[RockDawg]

OK, this is odd.  I haven't actually looked at the cache usage since right after copying everything back to it when it was showing 211Gb used.  It is now showing 129GB used.  I haven't yet tried JorgeB's suggestion so that isn't it.  I guess all's well that ends well.

 

So I guess my last question is why did my custom Docker network disappear.  Is that expected behavior or did I do something wrong that caused that?

[trurl]

1 hour ago, RockDawg said:

So I guess my last question is why did my custom Docker network disappear.  Is that expected behavior or did I do something wrong that caused that?

As I discovered custom docker network is apparently kept in docker.img, so when you recreate docker.img you must recreate custom network. It is a simple one line command best done before reinstalling containers.

[RockDawg]

Thanks!  Good to know it wasn't my mistake and it is easy enough to recreate but it does create confusion until you figure what the issue is..  I initially tried to copy my docker.img to the new cached drive but Unraid tols me I needed to rebuild it.  Any reason why it didn;t like the old one copied back over?