RAID-Z in a disk-failure.
Alexander Leidinger
Alexander at Leidinger.net
Sat Mar 26 21:46:51 UTC 2011
On Sat, 26 Mar 2011 15:42:57 +0100 Anders Andersson
<pipatron at gmail.com> wrote:
> Evening! I have a question about ZFS in a FreeBSD setting, more
> specifically about RAID-Z. I have never used ZFS so I might have
> misunderstood something, but imagine this situation:
>
> You're using ZFS in RAID-Z with 4 disks. One of these gets a
> catastrophic failure. You replace the disk with a new one, but when
> the RAID-Z is rebuilt, the software notices that one sector/block on
> another disk has become corrupted. It notices this because ZFS keeps a
> checksum.
>
> What happens then? Since the redundancy is temporarily disabled
> because of the failed disk, this sector/block is nowhere to be found.
> My hope is that the system will handle this gracefully, so that only
> the file using this block will be unreadable, but the rest of the data
> is available and can be rebuilt. The worst that could happen is that
> the rebuild is refused and the whole pool is gone.
>
> Have I missed something in this scenario?
If a corruption can not be corrected, tt's only the file (or
directory-subtree) with this corruption which is unreadable.
Bye,
Alexander.
--
http://www.Leidinger.net Alexander @ Leidinger.net: PGP ID = B0063FE7
http://www.FreeBSD.org netchild @ FreeBSD.org : PGP ID = 72077137
More information about the freebsd-fs
mailing list