zfs destroy of snapshot: dataset is busy

[Berend de Boer via illumos-zfs]

Hi All,

I cannot destroy a snapshot:

zfs destroy u1/***@snapshot
cannot destroy snapshot u1/***@snapshot: dataset is busy

I'm afraid I have to reboot this machine to clear the problem. Any suggestions before doing that?

--
All the best,

Berend de Boer



-------------------------------------------
illumos-zfs
Archives: https://www.listbox.com/member/archive/182191/=now
RSS Feed: https://www.listbox.com/member/archive/rss/182191/23047029-187a0c8d
Modify Your Subscription: https://www.listbox.com/member/?member_id=23047029&id_secret=23047029-2e85923f
Powered by Listbox: http://www.listbox.com

[Jim Klimov via illumos-zfs]

Post by Berend de Boer via illumos-zfs
Hi All,
I'm afraid I have to reboot this machine to clear the problem. Any
suggestions before doing that?
--
All the best,
Berend de Boer
-------------------------------------------
illumos-zfs
Archives: https://www.listbox.com/member/archive/182191/=now
https://www.listbox.com/member/archive/rss/182191/22497542-d75cd9d9
https://www.listbox.com/member/?&
Powered by Listbox: http://www.listbox.com

This rings a couple of bells:
1) dataset really is used, i.e. as a target or source of zfs-send/zfs-recv. I think other reasonable causes like holds or child datasets have their very different messages.
2) in very old days (early opensolaris) this might pop up and mount/unmount of the related dataset might help.
3) short of a reboot, an export-import of the pool (if successful) should help. But beware that if this is some lockup in zfs itself, you'd better have a few extra shells pre-opened and a (remote) console ready. In particular, you might background an '(sleep 600; uname 1 2)&' as a cheap deadman timer before reimporting the pool. Just foreground and break that timer if all goes well.

Hth,
Jim
--
Typos courtesy of K-9 Mail on my Samsung Android

[Berend de Boer via illumos-zfs]

Jim> This rings a couple of bells: 1) dataset really is used,
Jim> i.e. as a target or source of zfs-send/zfs-recv. I think
Jim> other reasonable causes like holds or child datasets have
Jim> their very different messages.

Well, it is used in the sense that the pool is sent to another system,
then a request is made to delete the snapshot.

Yeah, there are no holds.


Jim> 3) short of a reboot, an export-import of the pool (if
Jim> successful) should help.

Ah, good point. May give that a try!

--
All the best,

Berend de Boer




-------------------------------------------
illumos-zfs
Archives: https://www.listbox.com/member/archive/182191/=now
RSS Feed: https://www.listbox.com/member/archive/rss/182191/23047029-187a0c8d
Modify Your Subscription: https://www.listbox.com/member/?member_id=23047029&id_secret=23047029-2e85923f
Powered by Listbox: http://www.listbox.com