[OpenIndiana-discuss] Zpool replacing forever

[Gary Mills]

On Sun, Jan 20, 2019 at 08:16:44PM +0000, Andrew Gabriel wrote:
> 
> zfs is keeping the old disk around as a ghost, in case you can put it 
> back in and zfs can find good copies of the corrupt data on it during 
> the resilver. It will stay in the zpool until there's a clean scrub, 
> after which zfs will collapse out the replacing-0 vdev. (In your case, 
> you know there is no copy of the bad data so this won't help, but in 
> general it can.)

I see.  That's a good explanation, one that I didn't see anywhere
else.  I suppose that the man page for `zpool replace' should advise
you to correct all errors before using the command.  That way, the
confused status I saw would not arise.

> So to fix, delete the corrupt files, and any snapshots they're in.
> Then run a scrub.
> When the scrub finishes, ZFS will collapse out the replacing-0 vdev.

That did work.  Thanks for the good advice.  I found that many of the
snapshots could not be removed because they had dependant clones.  In
each case, it was a BE that was the dependant clone.  When I removed
the BE with beadm and did a scrub, the errors moved to another
snapshot.  I wound up deleting all the previous BEs before I got a
clean status after a scrub.  I suppose I could have found another way
to deal with the snapshots, but I didn't want to interfere with what
beadm expected.  I did do a `zpool clear rpool' at the end.  Here's
the status now:

            NAME          STATE     READ WRITE CKSUM
            rpool         ONLINE       0     0     0
              mirror-0    ONLINE       0     0     0
                c5t0d0s0  ONLINE       0     0     0
                c5t1d0s0  ONLINE       0     0     0
    
    errors: No known data errors



-- 
-Gary Mills-		-refurb-		-Winnipeg, Manitoba, Canada-