Why would doing zpool destroy
come back with “pool is busy” when there is nothing mounted and no scrub operation. I’d like to blast the pool and start over (renaming it) but nothing I do seems to work. I’ve run iostat on the device and nothing is accessing it.
Any help appreciated. Running Debian 12 with zfs 2.1.11-1.
thanks!
edit: fixed post typo (thanks Jim!)
zfs destroy
destroys a dataset (including zvols and snapshots), not the entire pool. If you want to destroy the pool itself, that’s what zpool destroy
is for.
thanks Jim, my post was wrong I was doing zpool. I’ll fix my original post.
Can you export the pool? Can you import the pool? Is this a root pool?
If you don’t care about anything on a pool, and you can export it, just use wipefs -a
on the partitions and THEN the disks that the pool belonged to once it’s exported.
If you can’t export it, you may have to shut down your system, physically unplug the drives, and boot the system, then plug in the drives, then do that.
If it’s a root pool, well, you’ve got to figure out what you’re doing with everything else first.
It was a regular pool (not a root pool), and I ended up using wipefs -a
.
There is one thing I didn’t mention: this was an external USB drive that I was using for snapshots. Could that have affected it? I suspect there’s a high probability.
I ended up removing the drive and connecting it via SATA to my other identical drive to create a mirrored pool. However, something strange occurred. With these 10TB drives, the resilvering process began as expected yesterday at 5 p.m., with an estimated completion time of 9 hours, which seemed reasonable.
When I checked last night, it was progressing normally. However, this morning, I noticed it appeared to have restarted the resilvering process, as it was only at 20%. There were no read or write errors. Is this normal behavior?
thanks!