zfs corruption after controller failure
InterNetX - Juergen Gotteswinter
jg at internetx.com
Tue Jan 6 15:29:45 UTC 2015
-T is an undocumentated and so for it should be considered as unstable
or even something which can easily break things. theres a reason why its
not documentated very likely.
dl;dr - its something i whouldnt touch...
What Output do you get from a simple zpool import ? You did everything
you tried until now with "readonly=on" ?
If everything seems to be broken, you could give the zfs forensic
scripts a try... its already broken, cant get worse anymore.
Am 06.01.2015 um 16:11 schrieb Fervent Dissent:
> I tried these, and some variations of them, and it gave the error the
> pool does not exist.
> zpool import -N -o readonly=on -f -R /pool <pool>
> zpool import -N -o readonly=on -f -R /pool -F -T <transaction_id> <pool>
>
> On Tue, Jan 6, 2015 at 11:08 PM, InterNetX - Juergen Gotteswinter
> <jg at internetx.com <mailto:jg at internetx.com>> wrote:
>
> > external usb drive (for sure cheap consumer desktop one)
> > cheap (maybe even buggy) usb controller
> > non-mirrored / raidz
> > probably cheap hardware at all (no ecc, like rainer mentioned for example)
>
> what do you expect. zfs is very tolerant, but some basics should be
> taken care of...
>
> but even with a single disk it should be possible to recover / import
> the pool at least read only
>
>
>
> Am 06.01.2015 um 16:06 schrieb Fervent Dissent:
> > I've had power lose before and a previous bad controller cause
> multiple
> > problems. The drive would disappear or go offline. I would clear
> it and go
> > on no problem. This is the first failure that I have not been able to
> > recover from.
> >
> > On Tue, Jan 6, 2015 at 11:01 PM, Rainer Duffner
> <rainer at ultra-secure.de <mailto:rainer at ultra-secure.de>>
> > wrote:
> >
> >>
> >>> Am 06.01.2015 um 15:58 schrieb Fervent Dissent <
> >> walkerindarkness at gmail.com <mailto:walkerindarkness at gmail.com>>:
> >>>
> >>> I have a external disk that was on a cheap usb controller, that
> >> controller
> >>> died.
> >>
> >>
> >> Maybe I’m mistaken, but I though that if your pool only has a
> single disk
> >> and that disk/pool shows errors or becomes
> unreadable/corrupted/whatever
> >> you cannot recover it.
> >>
> >> Same for not using ECC memory…
> >>
> >>
> >>
> >>
> > _______________________________________________
> > freebsd-fs at freebsd.org <mailto:freebsd-fs at freebsd.org> mailing list
> > http://lists.freebsd.org/mailman/listinfo/freebsd-fs
> > To unsubscribe, send any mail to
> "freebsd-fs-unsubscribe at freebsd.org
> <mailto:freebsd-fs-unsubscribe at freebsd.org>"
> >
>
>
More information about the freebsd-fs
mailing list