[OpenIndiana-discuss] Help debugging and replacing failed (maybe) hard drive

[Wood Peter]

>
>
> > - If the drive has to be replaced could somebody please confirm that the
> > following steps are sufficient:
> >
> > * zpool offline pool01 c3t0d0
> > * zpool detach pool01 c3t0d0
> > * Physically replace the drive with a new one
> > * zpool add pool01 spare c3t0d0
> > * zpool scrub pool01
>
> I would do this a bit differently:
> * zpool offline pool01 c3t0d0
> ** IF you believe the hard drive is faulty (vs. random single error),
> physically replace
> * zpool replace pool01 c3t0d0
> ** wait for automagic resilver (~36 minutes)
> * zpool detach pool01 c3t14d0
> * zpool add pool01 spare c3t14d0
>
> I like "my" way better because the drives/spares stay in the same
> locations (you might have them labelled a certain way) but it does
> require an additional copying of the data back (if you left checksums
> on, this shouldn't be a problem). No additional scrub should be
> necessary, but if you are doubting the drive, you can re-scrub just in
> case (you should be scrubbing weekly or monthly anyway, right?).
>
>
Thanks Jan. I like "your" way too. :-) I guess the wait and the extra
copying of the data is worth doing to have consistency in the physical
location of the spare drive across the storage servers.

Thank you all that replied. Very helpful information.

--
Peter