Brief description of the problem
Hi again. I’m replacing another disk (4TB) with a larger one (8TB) as it has become available.
Over 90 minutes after the rebalance started, nothing seems to be happening. I know it’s going to take a while, because I’ve already replaced one.
Anyway, nearly 2 hours after starting, 0% progress, and the new drive is still showing 1GB allocated. The new pool size is showing correctly though.
Detailed step by step instructions to reproduce the problem
- While shutdown, added ZA1FZPBY physically to /dev/sdc (having moved the 4TB drive it will replace (Z306JZ39 ) to /dev/sde.
- Wiped the new disk, as last time, but this time on the Rockstor box, not my ubuntu desktop. (fdisk to remove the partitions, then wipe the disk in the Assign role page, from Disks)
- In Pools, Resize/Re-raid, Add disk, select the new drive and Submit
(The last time, I added the 8TB (WKD391VC) to /dev/sde first, and after removing the 4TB it replaced, and rebalanced, shutdown and moved the new drive to the old drive’s position. From this I figured that the drives are identified by their serial number, not which file in /dev they are mapped to… maybe I was wrong… but, everything started fine before adding the disk to the pool.)
See both the existing 8TB drive (WKD391VC, not the one being replaced, but the previous one added) and the new one ( ZA1FZPBY ) are showing unknown power status.
New disk still only has 1 GB allocated.
Balance is running, still 0% progress
See how quiet the disks are. I would expect a flurry of writing going on.
ssh’d in. btrfs fi us shows the new drive, /dev/sdc, has no data, and 1GB of metadata, (the other 8TB drive, /dev/sdd, has 8GB of metadata)
Existing data seems not to have budged a bit.
Error Traceback provided on the Web-UI
What’s the best way to deal with a hung re-balance?
(Fortunately I have backup copies of the data…)
I’ve tried a btrfs balance pause, and that just hung.