i have a problem with IBM x3650 M3 raid 50 with 16 sas 146gb disks. Before christmas one of them started to lit yellow, but the machine wasworking (running esx). I have oredered a new one, but before i managed to replace it another one died, and the esx does not boot any more. Now i am not sure if the system will be able to rebuild the drive. The drive is divided into two spans, and both damaged disks are from the same span. Can anyone advise me what steps should i perform now before repleacing the disks?

  • 1
    Obligatory: RAID is not backup, it's just a way to minimize downtime. If you're using RAID 5 or any variant of it, you must have spares on hand unless recreating the array is nearly painless. Why do we keep seeing questions like this? Jan 3, 2017 at 12:15
  • 1
    @emtezet If both failed drives are in the same span, you likely cannot rebuild any more and need to replace the drives, reinitialize the array and restore from backup.
    – Sven
    Jan 3, 2017 at 12:22
  • Wow, that's a bad, because previous administrator did not bother in making backups, or having spares i started working here just before christmass... what about if i would be able to repair one of the actual disks (assuming it is a hardware issue on the elecrical board)? The virtual drive is now marked as offline. Would be possible to make it alive without loosing the data?
    – emtezet
    Jan 3, 2017 at 12:36
  • @emtezet: You sound like you are in way over your head. Don't mess with the actual drives. Follow @shodanshoks advice regarding ddrescue and don't touch the broken drives for anything else. The more you mess with them, the less the chance for data recovery firms to fix things if you decide to go that route in the end.
    – Sven
    Jan 3, 2017 at 16:58

1 Answer 1


If both failed disks are from the same span you are screwed, as it means a dual failure in a RAID5 subarray.

If you don't have/can't restore from backups and the failed disks are not completely broken, the only thing you can do before contacting a data recovery firm is to order two new SAS drives and use ddrescue or testdisk to recover as much data as possible from the failed devices. If they went bad only due to some bad sectors, and you are very lucky, maybe you can revive the entire array.

If all else fails, you must contact a professional data recovery service.

  • Thankyou for your anwser. Considering that i would be able (and lucky) to restore the data and put on a new disk will the array try to run itself or shoud i try to restore it with the megaraid tools? In the second case what steps will lead me to most harmless solution?
    – user393223
    Jan 3, 2017 at 16:26
  • After ddrescue, the new drivers should be automatically detected by the RAID controller. Anyway, after imaging the old disk, don't mess with them anymore. If you are uncomfortable with disk imaging, please stop here and contact a professional recovery service.
    – shodanshok
    Jan 3, 2017 at 17:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .