Thanks for the $0.02 -- Savannah's SCSI cabling is set up the same as it has been for three years. The three disks that failed were on separate SCSI buses. When we brought it in for servicing, we redid all of the cabling and reseated the SCSI controller. It was after this point that the fourth disk failed.
> [beuc - Fri Mar 16 15:14:20 2007]: > > Hi David, > > I forward you mail to the FSF sysadmins who were physically in front > of the mahcine :) > > On Fri, Mar 16, 2007 at 11:47:04AM +0100, David Kastrup wrote: > > > > Hi, > > > > I just had some association in connection with the massive RAID > > failure (several old and a new SCSI disk) on Savannah: while you > > probably have already checked for this: > > > > this sort of phenomenon strongly reeks of a problem with termination > > and/or cabling. Particularly since you report _one_ _new_ _disk_, it > > is quite likely that the corresponding change is responsible for the > > problem. > > > > Of course, this also implies a mechanical change with more > > possibilities for failures (controller no longer firm in slot, for > > example). > > > -- Joshua Ginsberg <[EMAIL PROTECTED]> Free Software Foundation - Senior Systems Administrator
