Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 13 Jul 2009 20:17:00 +0100 (BST)
From:      Chris Hedley <freebsd-current@chrishedley.com>
To:        current@freebsd.org
Subject:   ZFS pool corrupted on upgrade of -current
Message-ID:  <alpine.BSF.2.00.0907132009040.2027@teapot.cbhnet>

next in thread | raw e-mail | index | archive | help
I seem to have a bit of a problem; or rather an extremely large problem.

When updating FreeBSD from -current circa mid February to a -current as of 
a couple of days ago, I've just noticed some very serious problems with my 
ZFS pool.

I've been using a raidz2 pool spread across 8 sata discs, numbered ad0 - 
ad7.  It's been working fine across numerous system updates until now; it 
may have something to do with the device numbering changing (i.e. ad0, ad1 
etc becoming ad4, ad6 and so on) so perhaps I should have exercised more 
care, but my pool has become badly degraded: two discs are reporting data 
errors (though at exactly the same block numbers, and I doubt the discs 
became faulty at exactly the same time as my update) and two have fallen 
off the pool altogether: I've no idea what's happened to them or whether 
or not it's quietly using them.  They're still listed in /dev, just not in 
zpool status.

I've attempted to replace one of the discs reporting data errors with my 
hot spare but it only gets about 5MB into the resilvering process when it 
starts over, and it's done that repeatedly.  My other option was to use 
the hot spare as a simple UFS backup and copy as much data to it as 
possible, rebuilding the pool, but now it won't let go; so I now have an 
unstable pool, I can't rebuild it, and I have no means of backing it up. 
I even tried going back to my old February build, but it can no longer see 
the ZFS pool.  Can anybody recommend a solution...?

Chris.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.BSF.2.00.0907132009040.2027>