RAID 1: Mirrored Disk Array A mirrored disk array is composed of a set of two physical hard drives, each of which contains a full copy of all data sent to the logical drive that represents the array. This has a couple of advantages; first of all, any data stored on a RAID 1 array is completely and automatically backed up, and in the event of the failure of one drive, the other can be substituted without a hitch. Secondly, data can be read from both drives simultaneously, increasing the speed of data retrieval.
Fault tolerance is the cornerstone of RAID 1. In this configuration, two identical physical drives are used, with one drive mirroring the information on the other. A RAID 1 configuration is ideal for data redundancy, though storage is more costly as only 1/2 the total drive space of both hard drives is available.
Data writes take just as long as usual however. In the event one of the drives in the array fails, a new drive can be added, the array rebuilt, and the RAID controller will duplicate the information onto the new blank drive.
The disadvantage of RAID 1 is that unlike striping, a mirrored array can use only half of its total free space for storage, since one disk is an exact duplicate of the other.
RAID 0+1 Striped array with mirroring
This RAID level combines the best features of RAID 0 and 1. It requires a minimum of four physical drives to implement, so it is not cheap. Essentially, two pairs of striped drives are mirrored together to provide fault tolerance. The mirroring provides the fault tolerance, though if any drive is lost, it must be immediately replaced and the array rebuilt, since it cannot handle the loss of more than one drive.
RAID 0+1 does retain the inherent disadvantage of mirroring, however; effective disk space is halved since two of the four drives are exact duplicates of the other pair. Many other implementations of RAID exist, nearly all sharing one common factor: The expense and complexity of the hardware controllers required to implement them.
Intended for business use, these levels of RAID use the parity system as explained above to provide varying levels of fault tolerance. RAID solutions at this level generally come as an add-in controller card or a dedicated storage rack and are intended to work hand-in-hand with hot-swappable hard drive mountings. With this setup, any failed drives can be swapped out for new ones on the fly, and the missing data quickly restored by using the parity data.
Many setups will perform this operation automatically while still maintaining close to normal operation.
No comments:
Post a Comment