Overview
This section provides information about RAID storage subsystems and the IBM® Enterprise Storage Server®.
RAID stands for Redundant Array of Inexpensive Disks and provides a method of classifying the different methods of using multiple disks to increase availability. With RAID, multiple physical disks appear to the Content Manager OnDemand server as one logical disk. RAID carries out the concept of data striping by spreading data over multiple disks; a single file is segmented and stored on multiple disks. RAID carries out the concept of data mirroring by duplicating data from one disk to a second disk; a single file is stored twice, on two different disks. A failed disk still allows users to access data on the array, and a replacement disk or online spare can be recreated while the array is in use.
| RAID Level | Description | Protection | Performance |
|---|---|---|---|
| RAID 0 | Data striping on multiple disk drives. | Poor; single disk failure. | Best; read and write requests can be met by any disk. |
| RAID 1 | Disk mirroring. | Good; any disk can fail and data is still accessible. | Good; read request can be met by any disk. |
| RAID 3 | Disk striping with parity disk, using interleaved bytes. | Good; if any disk fails, data can be accessed by using information from other disks and parity disk. | Good for large data transfers. |
| RAID 4 | Disk striping with parity disk, using interleaved sectors. | Good; if any disk fails, data can be accessed by using information from other disks and parity disk. | Good for large data transfers. |
| RAID 5 | Disk striping with distributed parity data. | Good; if any disk fails, data can be accessed by using information from other disks and parity information. | Good for small block sizes. |
| RAID 5 Orthog –onal | Disk striping with distributed parity data, using dual controllers. | Best; if any disk fails, data can be accessed by using information from other disks and parity information, with additional protection from any single disk controller failure. | Good for small block sizes; improved performance because of use of dual controllers to read and write data. |
| RAID 6 | An extension of RAID 5 with a second portion of parity data to ensure disk recovery in the event of a second disk failure before the first drive can be fixed. | Provides independent access to the disks with floating parity to avoid the separate parity disk bottleneck of RAID 4. Multiple concurrent accesses to the array devices are supported to satisfy multiple concurrent I/O requests. | Data is striped in blocks across disk drives. Single records or tracks can be read from one disk drive without accessing other drives in the array. |