This is a "classical" problem on RAID5 arrays: when one disk die, you discover the others have developped bad sectors. Under this circumstance, many hardware RAID controllers simply drop the entire array (or refuse to start it).
If your RAID controller has a "forced" mode to bring the array online (albeit with some unavoidable data corruption), use it.
If not, you should be able to recover doing that:
- obtain a disk identical to the one with bad sectors
- connect the two disks to a Linux box with
ddrescue (or dd_rescue, in Debian parlance)
- copy the old disk to the new one using the command
ddrescue <bad_drive> <new_drive>
- let it finish and then connect the new drive to your RAID controller, in the same slot/position assigned to the bad one
- your array should go online, albeit with some unavoidable data corruption.
- do a complete filesystem check on the vdisk/array
If you are lucky, the corrupted data reside in the free/unallocated space, resulting in no data loss.
Otherwise, you risk losing something. If this is an unacceptable risk, forgot the step above: destroy/recreate your array and restore from backups.