Victor Latushkin wrote:
This issue (and previous one reported by Tom) has got some publicity
recently - see here
http://www.uknof.org.uk/uknof13/Bird-Redux.pdf
So i feel like i need to provide a little bit more information about the
outcome (sorry that it is delayed and not as full as previous one).
Morning,
Right, the PDF on there doesn't really give the full story of the
presentation, unfortunate as I see it seems to have got around a bit.
In the actual presentation I wasn't perhaps as harsh as it seems on the
slides!
First permanent error means that root block of the filesystem named
'content' was corrupted (all copies), so it was not possible to open it
and access any content of that filesystem.
Fortunately enough, there were not too much activity on the pool, so we
decided to try previous states of the pool. I do not remember exact txg
number we tried, but it was something like hundred txg back or so. We
checked it with zdb and discovered that that state was more or less good
- at least filesystem content was openable and it was possible to access
its content, so we decided to reactivate that previous state. Pool
imported fine and contents of 'content' was there. Subsequent scrub did
find some errors but I do not remember exactly how much. Tom may have
exact number.
I can't remember how many errors the check found, however all the data
copied off successfully, as far as we know.
--
Tom
// www.portfast.co.uk -- internet services and consultancy
// hosting from 1.65 per domain
_______________________________________________
zfs-discuss mailing list
zfs-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
