Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 26 Oct 2015 16:00:23 +0100
From:      Niklaas Baudet von Gersdorff <niklaas@kulturflatrate.net>
To:        Matthew Seaman <m.seaman@infracaninophile.co.uk>, freebsd-questions@freebsd.org
Subject:   Re: cd /.zfs/snapshot hangs (tmux put to uninterruptible sleep)
Message-ID:  <562E4007.5030206@kulturflatrate.net>
In-Reply-To: <562E1219.4080203@infracaninophile.co.uk>
References:  <562CB2C2.6090402@kulturflatrate.net> <562CB8CF.7010504@FreeBSD.org> <562CC1DD.9060207@kulturflatrate.net> <562CC308.6060901@FreeBSD.org> <562CC547.8080308@kulturflatrate.net> <562CC7B6.5040404@FreeBSD.org> <562CC939.5060807@kulturflatrate.net> <562E0DDA.3020202@freebsd.org> <562E1219.4080203@infracaninophile.co.uk>

next in thread | previous in thread | raw e-mail | index | archive | help
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 26/10/15 12:44, Matthew Seaman wrote:
> Actually, here is the problem:
> 
> stingray:/:# zfs destroy -r tank/.......1 cannot unmount
> '/.......1': Device busy
> 
> But that is easily fixed by:
> 
> stingray:/:# zfs umount -f /.......1 stingray:/:# zfs destroy -r
> tank/.......1
> 
> I wonder if that's worth a PR?  I recall now this is exactly what I
> ran into before, but it seems rather different to what you're
> seeing.

Aha!

I am not sure but it could be that it is related. (Sorry, unlike you I
haven't further investigated on what caused the error i.e., I am not
100% sure.)

What I realised yesterday was that I had several more processes stuck
in D. All of these were caused by a script [1] hat I use to create
snapshots of my ZFS filesystems regulary. It creates snapshots and
*deletes* old ones.

1: https://bitbucket.org/mmichele/zfssnap

My guess is now that the script caused the uninterruptible sleep of a
process (several times because I run it quite regularly) when it tried
to destroy an old snapshot. Then I came in and (because I wanted to
restore an accidentally deleted file) `cd /.zfs./snapshot` and
couldn't get access (the process hang) because the script already
messed up everything.

Does this make sense?

- -- 
Niklaas
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
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=6NuN
-----END PGP SIGNATURE-----



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?562E4007.5030206>