RAID (redundant array of inexpensive disks) is the technology of grouping several hard disk drives in a server into an array that can be defined as a single logical drive. This logical drive then appears to the operating system as a single physical drive. This technology provides access and data transfer rates beyond the physical limitations of existing hard disk drives, greatly enhancing logical-drive capacity and performance. In addition, if one of the physical drives fails, the system continues to run, with no operator intervention required. The defunct drive can be replaced without turning off the server (hot-swap) and the new drive contents are rebuilt from the information on the other drives. This rebuilding process takes place in the background with the system online. Once the drive contents are rebuilt, full performance and fault-tolerant functions are restored.
When you install internal SCSI drives in your server, you can connect them to either the Channel 1 or the Channel 2 connector on the RAID adapter (see the User's Handbook for more information). The external connector on the adapter also is designated "Channel 2". You cannot connect devices to both the internal Channel 2 and external Channel 2. This means that you can connect SCSI devices to both the internal channels and not the external channel; or, you can connect SCSI devices to one of the internal channels and the external channel, but never all three channels.
Disk Drive Capacities
Hard Disk Drive Mapping
One Channel Mapping
Two Internal Channels Mapping
One Internal and One External Channel Mapping
If you install another IBM SCSI-2 Fast/Wide Streaming-RAID Adapter/A in your server, you can use the same combinations shown here, depending upon your space availability.
When you create an array, you are combining several hard disk drives into one storage area. The array then can be used as a single logical drive or can be subdivided into several logical drives. A logical drive of a disk array can be any size you choose within the size limitations of the array. The RAID Adapter supports up to four independent arrays and a total of eight logical drives. Each array can be formed from a maximum of eight drives. And an array can span both channels. (For information about physical drives supported, see Data Storage Devices.)
For example, if you have only one array, it can be either a single logical drive or divided into as many as eight. If you have two or more arrays, you can have each one as one logical drive (a total of four), or you can divide them into multiple logical drives, as long as the total number of drives for the arrays is no more than eight.
The operating system considers each of these logical drives just as it does a physical hard disk drive. That is, the logical drives can be partitioned by the FDISK program (or its equivalent) in the same way that the operating system partitions a hard disk drive.
If you install an operating system, you either can allow the installation program to determine how the FDISK program (or its equivalent) allots the space within a logical drive, or manipulate the FDISK program yourself to partition the available space. The documentation you receive with your operating system explains how it handles mapping.
The ability to respond to multiple data requests provides not only an impressive increase in throughput, but a decrease in response time. The combination of parallel transfers and simultaneous responses to multiple requests allows disk arrays to provide the highest level of performance in network environments.