Markus Kilbinger writes:
> >>>>> "Greg" == Greg Oster <oster@cs.usask.ca> writes:
> 
>     Greg> In the first printout, the component labels are printed as
>     Greg> the RAID set is configured. At that point, even though the
>     Greg> raw partition of the RAID set is open, none of the component
>     Greg> labels will have been have been updated to say they are not
>     Greg> clean (configuration is a special case).
> 
>     Greg> The 'raidctl -s raid0' always shows what the label is
>     Greg> currently set to as well, but will always say "Clean: No",
>     Greg> given that it has to open the raw RAID device in order to
>     Greg> read the component labels.
> 
>     Greg> So the 'problem' here is that the information is being
>     Greg> obtained at different times, and while accurate, is somewhat
>     Greg> confusing/contradictory :(
> 
> It could not figure this out by myself. Thanks for your explanation.
> -> Shouldn't the second "Clean: No" changed in something like "In use"
> or "Busy"?

Removing it altogether would be even better (since it doesn't provide any 
useful info at that point).

>     >> Is anything wrong with the raid?
> 
>     Greg> No. "Parity status:" says 'clean', so things are fine.
> 
> That's the good news! :-) The problem was simply whom to believe...
> 
> Markus.

Later...

Greg Oster