Rebuild any RAID array other than 0, (because 0 is built for speed not redundancy). If one or more hard disks have failed, you can rebuild the array quickly and with no loss of data. This is relatively easy to do, since it is built to accept replacement hard disks without stopping. Do this as quickly as possible, since the likelihood of more than one failed volume increases the risk of data loss.
Using ServeRAID Manager Interface
Open ServeRAID Manager interface and find that the physical drives are listed to the left and the logical volumes are on the right. You will clearly see how many drives are attached on the left and if the device is networked storage, the controllers are also listed there. The software is intuitive inasmuch as if you want to view an item, just click it and it will open up another more detailed view. The view allows monitoring of the following items: fan, battery, temperature and power modules. If there is a risk involved, it will show the risk in a color code: blue is normal; yellow is a warning and red means failed. In "parity" arrays like RAID-5 and RAID-3, a failed drive's performance will suffer.
Rebuilding the Volume
The controller will normally detect the removal of a disk and also the replacement of a new drive when it is a part of an array. Go into the physical view (left-hand side) and click "Rebuild" on the drive that has been replaced. You can also force the rebuild of a drive that has gone bad but that does not necessarily require replacing. Do this by clicking "Rebuild" and the process will start as if the drive had been removed and replaced. In many systems, the spare will be already "hot" and will therefore begin the rebuild process before a request is made to "Rebuild" because the physical drive is already in the machine. The rebuild process may take hours or even days, depending on whether it is a parity RAID array, how much data has to be rebuilt and, of course, the size of the array. Also, if the system has kept a log of all the changed blocks, this will help speed the rebuild.
Best Practices for Drive Rebuilds
It is a wise idea to keep a set of the exact same drives on hand as the ones in the array. A well-functioning RAID array will come to a point when, when you least expect it, a drive will fail. If you do not have the luxury of hot-swapping drives already in place or you have allowed the drives that serve as back-up to your running drives to go bad without replacing them, it would be prudent to acquire some identical drives for the next time they fail. As a rule, a drive working 24/7 in a server, will have a duty cycle of around three years. In the unfortunate situation where a drive fails and you need to recover data from it, you can attempt to recover data from a failed Level 5 or 0 RAID Array using some software from Runtime (see link in Resources). This has more than a better chance of success and if data is crucial, this would be a great investment.
Tags: RAID array, data from, drive that, hard disks, more than
