Let me start by saying that I am new to Unraid but not to Linux/Unix. I have been using *nix based systems in one form or another for longer than I care to think about. Having said that I'm new to Unraid and ZFS and I'm worried that I'm doing something stupid or that I just don't understand something fundamental. I've tried searching the forums and of course I've tried google, but I can't find anything that is related.

This is a new install that I am testing. I've created a number of shares during my testing and I want to delete a few. For this test I have created two new shares that I want to delete: Test1 and Test2. Both are empty and have never contained any data. Both are configured identically except that Test2 is exported to NFS.

Test 1 is identical, but I've set the export flag to no and applied the change. If I delete Test1, delete process completes exactly as expected and I'm returned to the normal UI in a couple of seconds. When I try to delete Test2 I get the dancing icon and it never stops:

I've let the system sit for an hour. Once this occurs, I'm stuck. I can't stop the array, I hit the stop button and nothing happens. I can't delete other shares, they all hang in the same fashion. I can't do anything. I can't even shut the system down cleanly. if I hit shutdown or reboot the system will get to the generating diagnostics line and hang. I can still interact with other shells I have open, but it will not shutdown. I'm typically forced to power the machine off.

Jan 23 10:06:37 hobbiton rc.0: Running shutdown script /etc/rc.d/rc.0:
Jan 23 10:06:37 hobbiton init: Trying to re-exec init
Jan 23 10:06:37 hobbiton rc.0: Saving system time to the hardware clock (UTC).
Jan 23 10:06:37 hobbiton rc.0: Creating system time correction file /etc/adjtime.
Jan 23 10:06:37 hobbiton rc.0: /sbin/hwclock --utc --systohc
Jan 23 10:06:38 hobbiton rc.0: Save PCI Configuration.
Jan 23 10:06:41 hobbiton rc.local_shutdown: Waiting up to 90 seconds for graceful shutdown...
Jan 23 10:08:11 hobbiton rc.local_shutdown: Forcing shutdown...
Jan 23 10:08:11 hobbiton rc.local_shutdown: Status of all loop devices
Jan 23 10:08:11 hobbiton rc.local_shutdown: losetup -a
Jan 23 10:08:11 hobbiton rc.local_shutdown: Active pids left on /mnt/*
Jan 23 10:08:11 hobbiton rc.local_shutdown: fuser -mv /mnt/cache /mnt/datapool /mnt/user
Jan 23 10:08:11 hobbiton rc.local_shutdown: Active pids left on /dev/md*
Jan 23 10:08:11 hobbiton rc.local_shutdown: fuser -mv /dev/md*
Jan 23 10:08:11 hobbiton rc.local_shutdown: Generating diagnostics...

So I've pretty much confirmed to myself that having the NFS share active is what is triggering this behavior, but I'm looking for thoughts from others. Is this an issue that is specific to the ZFS pool? Is this something with Unraid? Is this a known issue that I just failed to find while searching?

Thanks,

GR.

Will do some more testing tomorrow, but looks like you have found a bug

Thanks. I really appropriate it. I've been beating my head against the random hangs for a week. Never tied it back to the deleting shares until yesterday. At first I was convinced it was a hardware issue and switched hardware.

I can reproduce, seems like an issue unmounting a dataset when it's being exported over NFS, reported to LT, and it should be fixed for a future release. The workaround for now is to stop exporting the share over NFS before trying to delete it.

Renaming share that is exported also seems to trigger it.

Yeah, it should be the same issue.

This should be resolved in 7.3.0-beta.1