Upgrading Cache drive to larger drive - Is SpaceInvaderOne's tutorial still valid?

[bobalot]

I have an existing Samsung 830 SSD that I'm using as a cache drive.

 

I want to replace it with my Crucial P2 500Gb M.2 SSD. It is already in my unassigned devices (see attached image).

 

I found SpaceInvaderOne's video for doing this. It is from 2017. Is it still valid?

 

Should I just follow his instructions or do the more recent versions of Unraid require a different procedure?

[JorgeB]

I don't know which method he mentions in that video, but you can do a direct replacement for a single device "pool" as explained in the link below, note that you need to be on v6.10.x, this was broken on v6.9.x:

 

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=480419

 

[bobalot]

4 hours ago, JorgeB said:

I don't know which method he mentions in that video, but you can do a direct replacement for a single device "pool" as explained in the link below, note that you need to be on v6.10.x, this was broken on v6.9.x:

 

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=480419

 

Thanks for the quick response!

   

I do have Unraid V6.10.2.

 

I stopped the array, swapped out the 250GB SSD with my new 500GB one and started up the array and got the error "Unmountable: Invalid pool config" (see attachment #2).

 

 

I do have 2 slots in my pool. See attachment.

 

 

Edited June 13, 2022 by bobalot
Additional information.

[JorgeB]

Please post the diagnostics.

[trurl]

1 hour ago, bobalot said:

swapped out the 250GB SSD with my new 500GB one

 

The linked procedure says

On 7/18/2016 at 4:46 AM, JorgeB said:

-This procedure assumes you have enough ports to have both the old and new devices connected at the same time, if not you can use this procedure instead.

 

Did you have both connected at the same time? Or did you use that other procedure linked there?

 

[bobalot]

1 hour ago, JorgeB said:

Please post the diagnostics.

Please find attached.

tower-diagnostics-20220613-2251.zip

[bobalot]

20 minutes ago, trurl said:

 

The linked procedure says

 

Did you have both connected at the same time? Or did you use that other procedure linked there?

 

I have them both connected. 

 

One is in my cache pool and the other is an unassigned device.

[JorgeB]

Hmm, will need to investigate, possibly it's not working again, for now do this:

 

-unassign all cache devices

-start array

-stop array

-in the console type:

btrfs-select-super -s 1 /dev/sdc1

If you rebooted since the diags check old SSD is still sdc

 

-if the command completes without an error re-assign old cache device and start the array

-cache should now mount

 

To replace for now you can add the new device as cache2, once the balance finishes, stop the array, unassign cache1 and start array again to remove it from the pool.

[bobalot]

I had no problems when reassigning my old ssd when I tried to swap the drives the first time. I just stopped the array, reassigned it, restarted the array and everything seemed to work fine. 

 

Do I still need to do the first part? 

Edited June 13, 2022 by bobalot

[JorgeB]

9 hours ago, bobalot said:

I had no problems when reassigning my old ssd when I tried to swap the drives the first time. I just stopped the array, reassigned it, restarted the array and everything seemed to work fine. 

Are you sure? Pretty sure I remember seeing the device being wiped, but it won't hurt to try, if it doesn't mount follow the instructions above, just make sure you don't do anything else to that device (just trying to mount it is not a problem).

[bobalot]

5 hours ago, JorgeB said:

Are you sure? Pretty sure I remember seeing the device being wiped, but it won't hurt to try, if it doesn't mount follow the instructions above, just make sure you don't do anything else to that device (just trying to mount it is not a problem).

No, you are right. Now the 830 SSD has the "Unmountable: Invalid pool config" error.

 

I have followed your instructions. Both devices are now unassigned.

 

I'm having problems stopping the array. The array never stops. It just says "Array Stopping: Retry unmounting disk share(s)".

 

Ignore the striked out. I rebooted and the array started up again and then I was able to stop it. 

 

I mounted my original SSD and added my new SSD to the pool as cache 2.

 

There is now a message saying:

 

"Unraid Cache disk message: 14-06-2022 22:38

Warning [TOWER] - pool BTRFS too many profiles (You can ignore this warning when a pool balance operation is in progress)
SAMSUNG_SSD_830_Series_S0Z4NEAC832198 (sdc)"

 

I'm guessing it is pool balancing now? I'll wait until it is over and then try removing the 250GB SSD. When do you when it is over?

Edited June 14, 2022 by bobalot

[JorgeB]

1 hour ago, bobalot said:

You can ignore this warning when a pool balance operation is in progress

It's normal because of the balance, once it finishes you can remove the other device.

[JorgeB]

1 hour ago, bobalot said:

When do you when it is over?

Stop array button should be inhibited while it's running.

[bobalot]

8 hours ago, JorgeB said:

It's normal because of the balance, once it finishes you can remove the other device.

I waited until the stop array button was available. Stopped the array and removed my old SSD and restarted the array.

 

I now have the "Unmountable: Invalid pool config" error.

 

In my Fix Common Problems settings tab, I have these errors (see attached):

 

WARNINGS FOUND: "Share XXXXXX references non existent pool cache" SUGGESTED FIX:"If you have renamed a pool this will have to be adjusted in the share's settings"

 

I have attached a diagnostics download.

 

BTW: Thanks for all your help with this. It is much appreciated.

tower-diagnostics-20220615-0743.zip

Edited June 14, 2022 by bobalot

[JorgeB]

Jun 15 07:25:22 Tower emhttpd: shcmd (485): /sbin/wipefs -a /dev/sdc1
Jun 15 07:25:22 Tower root: wipefs: error: /dev/sdc1: probing initialization failed: Device or resource busy

 

Strange, for some reason device failed to wipe due to being busy, with the array still running see if you can do it manually now, by typing:

 

wipefs -a /dev/sdc1

 

Then stop the array and start again, if it still shows busy disconnect the device physically and start the array.

[bobalot]

4 hours ago, JorgeB said:

Jun 15 07:25:22 Tower emhttpd: shcmd (485): /sbin/wipefs -a /dev/sdc1
Jun 15 07:25:22 Tower root: wipefs: error: /dev/sdc1: probing initialization failed: Device or resource busy

 

Strange, for some reason device failed to wipe due to being busy, with the array still running see if you can do it manually now, by typing:

 

wipefs -a /dev/sdc1

 

Then stop the array and start again, if it still shows busy disconnect the device physically and start the array.

I tried typing the following into the console:

 

wipefs -a /dec/sdc1

 

And it came up with the following error:

 

wipefs: error: /dev/sdc1: probing initialization failed: Device or resource busy

 

Edited June 15, 2022 by bobalot

[JorgeB]

Don't remember seeing that before, recommend shutting down, disconnecting that device and powering back on.

[bobalot]

1 hour ago, JorgeB said:

Don't remember seeing that before, recommend shutting down, disconnecting that device and powering back on.

 

Done. It seems to have fixed the issues in the "Fix Common Problems" settings page. I do get this now (see below). Looks a bit weird, no size or used space is stated.

 

 

Edited June 15, 2022 by bobalot

[JorgeB]

1 hour ago, bobalot said:

I do get this now (see below). Looks a bit weird, no size or used space is stated.

It is weird, there's no device on slot 1 but it should still show the usage graphs, could be related to the theme you're using, but just assign it to slot 1:

 

-stop array

-unassign the device from cache2

-re-assign to cache1

-start array

 

Note that if you ever want to use the old cache device in the same server you need to wipe it before array start, you can use:

blkdiscard -f /dev/sdX

Replace X with correct letter.

[bobalot]

11 minutes ago, JorgeB said:

It is weird, there's no device on slot 1 but it should still show the usage graphs, could be related to the theme you're using, but just assign it to slot 1:

 

-stop array

-unassign the device from cache2

-re-assign to cache1

-start array

 

Note that if you ever want to use the old cache device in the same server you need to wipe it before array start, you can use:

blkdiscard -f /dev/sdX

Replace X with correct letter.

It worked!!!

 

Thanks so much for your help!

[bobalot]

@JorgeBSorry, one last query.

 

When I started Sonarr, I got this error:

 

"database disk image is malformed"

 

I didn't do anything else to my server, so I'm assuming this is related to the SSD swap over.

 

I have restarted the docker and the array, and it didn't seem to fix the problem. 

 

Is there a way to fix this? 

[JorgeB]

Swap should not cause any issues, data is still the same, but can't really help with that, suggest you ask for help in the appropriate docker support thread, there's a likely a way to repair de db.

[bobalot]

1 hour ago, JorgeB said:

Swap should not cause any issues, data is still the same, but can't really help with that, suggest you ask for help in the appropriate docker support thread, there's a likely a way to repair de db.

Will do! 

 

Thanks for all your help!