HP Surestore Disk Array FC60 HP SureStore E Disk Array 12H User's and Service - Page 262

Glossary Glossary Auto Rebuild Availability Balancing Block Mirroring Cache Channel Configuration Data Redundancy Auto Rebuild begins immediately if a disk failure occurs, as long as enough space is available to perform the rebuild. No operator intervention is required to perform an Auto Rebuild. Once the Auto Rebuild has completed, the disk array is once again Fault-tolerant, since all user data is once again redundant. Availability is achieved by using redundant data to prevent the loss of use of a storage system in the event of a disk failure. Balancing is the process of automatically spreading data equally across all disks to increase performance. When a fixed amount of data is either written or read from multiple disks instead of to or from just one disk, the throughput or speed of the process increases greatly. Block Mirroring is a technique in which duplicate copies of blocks of data are stored on an array of disks. An on-board cache, or memory buffer, greatly enhances the speed of data transfers to and from disk devices, since the next block of data required by the host computer is often already available in high-speed cache memory. Caching does require controller overhead, however, so unlike traditional RAID systems that often utilize huge cache space, the disk array is tuned with an optimum cache. The disk array contains SIMMs with Error Correction Code (ECC), which are capable of both error detection and correction on the Disk Array Controller. Most RAID systems use SIMMs without ECC. A Channel refers to a SCSI bus on the Disk Array Controller. Each Disk Array Controller is connected to one channel. See Auto Configuration and Self Configuring. Data Redundancy protects the disk array from lost data if a disk fails. With Data Redundancy, the array can reconstruct the data that was on the failed disk. If a disk fails, Data Redundancy is lost until the array rebuilds the data that was on the failed disk. 262