Replace the hard drive with a drive that is the same type and the same size, or larger. It will rebuild itself.
By same type I mean that the connector on the back of the drive and the speed of the drive interface must be the same. If the new drive spins at a different rate it will not create a problem, except that if it is slower it will make read/writes slower. If it is larger it will only use part of its capacity, matching the capacity of the other drives.
Note that if you were to replace all drives with larger drives (one at a time) the capacity of the entire array would be the same as the original. If you want a larger array it must be built from the start with that larger capacity.
Take out the bad drive. If you can get the exact same drive, do it, otherwise look it up on the manufacturers web site to find out size and type.
While there are a few possible RAID configurations that you could be using on that server, the chances are overwhelming that it is RAID level 5 that you are using. This RAID level is tolerant of a single drive failure. If you lose another drive before your new drive rebuilds your data will be lost. Hopefully you have a backup.
RAID 5 stripes the data and the parity data across all available drives and when a drive fails, the missing data is rebuilt on the fly from the parity data. When a replacement drive is installed the parity data is used to rebuild the data that belongs on that drive. The advantages of RAID 5 are fault tolerance (of a single drive) and speed because each drive only has to provide or write its portion of the total data being accessed or written.
May 26, 2009 |
Acer Altos G710 Server