On Wed, 18 Jun 2003, Ernest E Vogelsinger wrote:
> At 18:12 18.06.2003, scott.marlowe said:
> --------------------[snip]--------------------
> >is worth it. Having lost a RAID5 set to a hw controller that simply had
> >the cable to two drives come loose but refused to accept them back into
> >the RAID5 after that without formatting them first, I'm no longer as wild
> >about hw raid controllers as I once was.
> --------------------[snip]--------------------
>
> Say - you didn't test it before going production?
> ;-))
Actually, that was a legacy Oracle box we had that problem on. And it was
someone else who called me RIGHT after moving it and having the cable come
loose. Asked where his backups were. "We don't need backups, we run on a
RAID5." uh huh... :-)
It was on the older MegaRAID 428, by the way.
> Never mind - I was that ignorant myself. However my current RAID V's are
> tested above the specs - installed RH7.2, removed 2 (sic!) disks, and
> remounted. The controller complained (of course) but still offered to try
> to remount - bingo, worked.
I test my linux box's sw raid the same way. Early flavors of sw raid were
a little goofy (RH6.2 for instance) but under 7.x they seem to be very
stable and work well at replacing drives and all.
> I feel somehow safe now.
Don't worry, something you forgot about will pop up it's head soon enough
;^)
