[zfs-discuss] Oops: Duplicate replace operation

Exec Unerd execunerd at gmail.com
Sat Feb 6 19:34:43 EST 2016


ZFS appears to use the last replace issued. The other device (e00p00d07) is
not attached to the pool.
          raidz2-11                 DEGRADED     0     0     0
            e00p00d00               ONLINE       0     0     0
            e00p00d01               ONLINE       0     0     0
            e00p00d02               ONLINE       0     0     0
            e00p00d09               ONLINE       0     0     0
            e00p00d04               ONLINE       0     0     0
            e00p01d00               ONLINE       0     0     0
            e00p01d01               ONLINE       0     0     0
            e00p01d02               ONLINE       0     0     0
            e00p01d03               ONLINE       0     0     0
            e00p01d04               ONLINE       0     0     0

It might be nice if ZFS would detect that a replace is already running and
at least warn you and require the -force option to override.


On Fri, Feb 5, 2016 at 7:05 PM, Exec Unerd <execunerd at gmail.com> wrote:

> I have a large pool of RZ2 raid sets, and a couple of drives failed today.
> No problem, as I have spares waiting, so I run the `zpool replace` command
> to spare-out the first drive. Then, while attempting to replace the second
> drive, I used the up arrow to recall my command and accidentally replaced
> the same drive twice.
>
> I don’t know what’s going to happen, but it seems to not be an error yet.
> Should I be concerned, or will ZFS just use the last drive it was told to
> resilver (e00p00d09)?
>           raidz2-11                 DEGRADED     0     0     0
>             e00p00d00               ONLINE       0     0     0
>             e00p00d01               ONLINE       0     0     0
>             e00p00d02               ONLINE       0     0     0
>             replacing-1             FAULTED      0     0     0
>               e00p00d03             FAULTED      0     0     0  too many
> errors
>               e00p00d07             ONLINE       0     0     0
>  (resilvering)
>               e00p00d09             ONLINE       0     0     0
>  (resilvering)
>             e00p00d04               ONLINE       0     0     0
>             e00p01d00               ONLINE       0     0     0
>             e00p01d01               ONLINE       0     0     0
>             e00p01d02               ONLINE       0     0     0
>             e00p01d03               ONLINE       0     0     0
>             e00p01d04               ONLINE       0     0     0
>
> I'm not panicking because it's a RAIDz2 and the other failed drive is in a
> different raid group (raidz2-13), so I'm going to wait it out for now,
> unless someone has stepped in this corner case before and has some idea
> what ZFS will do.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.zfsonlinux.org/pipermail/zfs-discuss/attachments/20160206/acff5919/attachment.html>


More information about the zfs-discuss mailing list