Re: raidframe query

[David Brownlee <abs%netbsd.org@localhost>]

On 21 September 2011 12:03, Patrick Welche <prlw1%cam.ac.uk@localhost> wrote:
> On Wed, Sep 21, 2011 at 11:47:33AM +0100, David Brownlee wrote:
>> You should be able to raidctl -R the failed unit to try to tell it to
>> rebuild, though I would be very suspicious of that error rate.
>
> I actually got it the wrong way around
>
> wd0a: optimal  wd0: 82 uncorrectable errors
> wd1a: failed   wd1:  0 uncorrectable errors
>
> raidctl -R /dev/wd1a raid0
> -> IO Error.  Marking /dev/wd1a as failed.
> -> initiating in-place reconstruction on column 1
> -> Recon write failed (status 30(0x1e)!
> -> reconstruction failed.
>
> raid1's wd0e / wd1e are both optimal
> raid2's wd0f / wd1f are both optimal
>
> I am not getting any failing I/O kernel messages from hptide which is what
> the drives are plugged in to. (The drives and hptide card were working in
> another -current/i386 box for many years as a raidframe mirror - I just
> wiped the drives and put the card+drives into another -current/i386 box
> on Saturday and I didn't drop them ;-) )

What does "dd if=/dev/wd1a of=/dev/null bs=1m count=1024" report, plus
is the disklabel for wd1 the same as wd0?