Avago Technologies 3ware 9650SE-12ML (Channel) - Manual (Page 180)

Maintaining Units
168 3ware SAS/SATA RAID Software User Guide, Version 9.5.1
A RAID 50 unit can sustain multiple drive failures, as long there is only one
failed drive in each RAID 5 set.
A RAID 6 unit can have two simultaneous drive failures, before becoming
inoperable.
When a RAID 5 or RAID 6 is running in Degraded mode and you rebuild it,
the missing data is reconstructed from all functioning drives.
Rebuilding a Unit Through 3DM
When a drive on a unit becomes defective, you replace it with an available
drive and then rebuild the unit. If Auto Rebuild is enabled and a drive is
available, starting the rebuild is optional as it will start automatically.
To rebuild a unit through 3DM
1 If necessary, add a new drive to replace the failed drive. (For details, see
“Adding a Drive” on page 142.)
2 In 3DM, choose Management > Maintenance .
3 In the Unit Maintenance section of the Maintenance page, select the
degraded unit and click the Rebuild Unit button.
4 When a dialog box displays available drives, select the drive you want to
replace the failed drive and click
OK .
Figure 67. Selecting a Drive when Rebuilding
5 If the degraded unit has more than one failed drive (for example, a RAID
10 where both mirrored pairs each have a failed drive), repeat step 3 and
step 4 to select another drive.
Note: If a rebuild fails, check the Alarms page for the reason. If there was an ECC
error on the source disk, you can force the rebuild to continue by checking the
Overwrite ECC policy on the Controller Settings page in 3DM and then running
Rebuild again. This will cause uncorrectable blocks to be rewritten and the data
may be incorrect. It is recommended that you execute a file system check when the
rebuild completes. Under Windows, you can do this by right-clicking on the Drive
and choosing
Properties ; then on the Tools tab, click Check Now . Under
Linux or FreeBSD use fsck /dev/sda1 .