April 12, 20242 yr Hi 🙂,  I'm fairly new to unraid and everything seemed to be working well, until now. I decided that I wanted to migrate my cache pool to a new drive, going from 128 GB SSD to 1 TB NVME. I've seen online that a simple way to do it would be to add a new drive to the cache pool, wait for BTRFS sync to complete, stop the array, remove the 128 GB and voila, the 1 TB should now be the default drive in the cache pool.  While the first few steps worked fine, I was indeed able to add the new drive to the cache pool, BTRFS synced happened, my array was then resized to 5XX GB (with 98% usage which seems suspicious), but I was able to stop the array, so I did, removed all drive from the cache pool, changed the slots to 1, added the 1 TB as the primary drive and started the array.  It started, however I quickly saw this message next to the drive : Cache drive unmountable: Unsupported or no file system  What are my options now? I haven't erased any drives and I've been trying to set the 128 GB back, or the 1TB, but haven't been successful. I included the diagnostics too! Thanks! crousti-server-diagnostics-20240412-1144.zip Edited April 12, 20242 yr by FrForget
April 12, 20242 yr Community Expert 6 minutes ago, FrForget said: removed all drive from the cache pool, changed the slots to 1, added the 1 TB as the primary drive and started the array. That won't work, to remove a device you just unassign the devices and leave the slots as they were, then start the array to begin the balance.  Stop the array, change slots to 2, start the array, if the pool doesn't mount post new diags.
April 12, 20242 yr Author Thank you for your answer. I just tried that and added the 1 TB back as the main cache pool drive. Unfortunately, I'm still seeing the same error : Then at the bottom, I also see this : Unmountable disk present: Cache • Samsung_SSD_980_PRO_1TB_S5P2NL0W409998F (nvme0n1) New diagnostics included too! crousti-server-diagnostics-20240412-1228.zip
April 12, 20242 yr Author There we go :Â Seems like 2 are found, is that right? However the sizes doesn't match. Edited April 12, 20242 yr by FrForget
April 12, 20242 yr Community Expert Solution Assuming that NVMe device should be the remaining one, type sfdisk /dev/nvme0n1 then type 2048 and hit enter, post the output of that
April 12, 20242 yr Author Here's the output : I haven't selected Yes or No. Too afraid to remove the data on the drive 🙃
April 12, 20242 yr Community Expert Type 'N' and enter to keep the signature, then 'write' and enter, then output of btrfs fi show Â
April 12, 20242 yr Author Alright, done! What's next 🙂? Going back to Main, I still see Unmountable. Do I need to Stop / Start the array? Reboot?Â
April 12, 20242 yr Community Expert Unassign the nvme device from the pool, start the array, stop the array, re-assign the device, start array, post new diags.
April 12, 20242 yr Author Sorry, I can't start the array until I select the "Yes, I want to do this" saying Start will remove the missing cache disk and then bring the array on-line.  That's what I want right ?Â
April 12, 20242 yr Author Well, I think we are up to something here : Interestingly it's showing 564 GB as the size? Added new diags as well. crousti-server-diagnostics-20240412-1353.zip
April 12, 20242 yr Author Interesting, now it says 1 TB! Now, previously, it was saying that a second drive was missing, does it now understand that only a single drive will remain in the cache?
April 12, 20242 yr Community Expert It's normal for the size stats to be wrong during the balance to single, just wait for the balance to finish, when ther pool activity stops, and you are done, you can then change slots to 1.
April 12, 20242 yr Author Looks like it is completed as I can stop the array and see the full 1 TB now.  I want to say a HUGE thank you for the instructions you've provided.  Overall, would you say that this is the proper way to do it? If I wouldn't have change the slots number to 1 and simply unassigned the second drive and wait for it to balance, I wouldn't have had to use the command line do fix the drive right?
April 12, 20242 yr Community Expert 5 minutes ago, FrForget said: Overall, would you say that this is the proper way to do it? If I wouldn't have change the slots number to 1 and simply unassigned the second drive and wait for it to balance, I wouldn't have had to use the command line do fix the drive right? Correct.
April 12, 20242 yr Author Again, thank you so much! The information you've provided here are really useful.
March 31, 20251 yr Hi all, I am having a very similar issue here but can't work out what I have done wrong.  I had a 128gb cache and am upgrading it to a 1tb.  I did the following in order: - stopped the array - added new NVME to cache - Started the aray and allowed btrfs jobs to finish - Stopped the array and removed cache 1 (the 128gb) - Started the array  After this i got the warning - Cache pool BTRFS missing device(s)No device identification ()  I then tried to reduce the number of cache drives to 1 but this wasnt an option.  Now I can't add the original back into the pool as i get an error when trying to start the array saying cache - replacement device too small.  I've attached the diagnostics so any assistance would be much appreciated. lawson-nas-diagnostics-20250331-2030.zip Edited March 31, 20251 yr by ldog88
April 1, 20251 yr Community Expert 11 hours ago, ldog88 said: I've attached the diagnostics so any assistance would be much appreciated. There's a bug with 7.0.x removing btrfs devices, post the output from btrfs fi show Â
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.