Needed to upgrade 2 of the 4 SSDs i had in my cache pool so i swapped in the first one and it did the standard BTRFS rebuild or whatever it is supposed to do. Once that completed i stopped the array and put the second one in, this time when it came up though it did a parity check for some reason and never actually did the BTRFS rebuild or whatever it is and now when i start the array it was telling me i had a missing cache drive but referenced the new drive as the missing one even though its in the array.

 

So my question is, what do i do so i can get the BTRFS function to do what it is supposed to do?

preferably with the array started
Go to Tools - Diagnostics and attach the complete Diagnostics ZIP file to your NEXT post in this thread.

@trurl attached

storage-diagnostics-20201221-0900.zip

 

Also, i stopped and started thee array before making the OP so i could get the exact message from the popup but it didnt come up this time. Below is a picture of my cache pool, sdc is the new drive that i put in that did not rebuild.

Edited December 21, 20205 yr by hermy65

Stop array, unassign cache1, start array, wait a couple of minutes and grab new diags.

@JorgeB attached

storage-diagnostics-20201221-1037.zip

Strange, it's not trying to remove the missing device, add it back again.

@JorgeB Done, attached are new diagnostics plus a screenshot of the popup i got. 

storage-diagnostics-20201221-1047.zip

Edited December 21, 20205 yr by hermy65

The notification is normal, just not normal that it's not trying to delete/replace the missing device, try this, stop the array, again unassign cache1, start array and type:

 

btrfs dev del missing /mnt/cache

Post new diags after cache pool activity stops.

 

 

 

@JorgeB Having issues getting this to work, keeps saying im out of space when there is almost 600GB available

storage-diagnostics-20201221-1351.zip

11 hours ago, hermy65 said:

when there is almost 600GB available

Stats with v6.8.x are wrong when using 3 devices in raid1, both used and free, cache is full, you need to free up some space then run device delete again.

@JorgeB Interesting. I just manually checked and im seeing about 335GB used out of 1TB. After that i moved another 70GB off of it so literally the only thing left on cache is my plex folder and my docker image which both of those total about 240. Im running the command now, will see what happens this time. 

@JorgeB You were correct as always sir, ran it again after moving that last 70GB and it looks like it completed. Diagnostics are attached. 

 

 

storage-diagnostics-20201222-0753.zip

OK, now add back the other SSD and this time it should hopefully work.

@JorgeB Is it normal for this to happen for that drive? It is a brand new drive.

No, don't format and post new diags.

@JorgeB attached

storage-diagnostics-20201222-0808.zip

Something very weird is going on with that pool, if you unassign cache1 again does it mount? If it doesn't post diags.

@JorgeB removed it and started it and looking in unassigned devices it doesnt show that its mountable, can preclear it though.

I meant does the pool mount without that SSD assigned.

@JorgeB yes the pool mounts, or at least appears to without that new disk in.  But if i look at the docker tab it tells me that docker isnt started, my vms arent available, etc

storage-diagnostics-20201222-0834.zip

Edited December 22, 20205 yr by hermy65

No, it's not mounting, and with an error that doesn't make sense:

 

super_num_devices 3 mismatch with num_devices 3 found here

Basically means it's expecting 3 devices and it's finding 3 devices, and that doesn't match.

 

Did you still have anything important on cache? Docker image can easily be recreated, as for Plex was that the database? I don't use Plex so not sure if that is easily recreated or if you have a backup.

 

@JorgeB just your standard vms, appdata, docker image, etc.

 

My question for you then is since the new SSD is still showing as unmountable in the unassigned devices but does give me the option to preclear it...should i do that? What is to say wiping out the cache and starting over would get that other one working?

 

Edit: Also, i still have the drive i pulled sitting in unassigned devices, if i put that back in will that break anything? Or is it better to put the new drive back in the array, wipe the cache array, then copy data from the old drive.

Edited December 22, 20205 yr by hermy65

@JorgeB also, i assume if we blow this away i just add the new drive then tell it to format the cache array since its unmountable or whats the correct process here?

41 minutes ago, hermy65 said:

Also, i still have the drive i pulled sitting in unassigned devices,

Do you mean the original device? If it's the original device try adding it back, but you need to this:

 

Stop the array, unassign all cache devices, start array to make Unraid "forget" current cache config, stop array, reassign all 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), start array.

 

1 minute ago, hermy65 said:

i assume if we blow this away i just add the new drive then tell it to format the cache array since its unmountable

Just need to format with the new device, all data will be lost.

 

 

 

@JorgeB bummer, that did not work. Followed your steps, there was no existing data overwritten warning. Started it back up and it says unmountable for the cache array