HP 166207-B21 Compaq RAID Advanced Data Guarding: A Cost-effective, Fault-tole - Page 3

TECHNOLOGY BRIEF Compaq RAID Advanced Data Guarding RAID - Redundant array of independent disks. Striping - The spreading of data over multiple disk drives to improve performance. Data is interleaved by bytes or by sectors across the drives. TC020604TB ... INTRODUCTION Customers are being flooded by a rising tide of data that is the result of e-business and traditional applications: transaction processing, enterprise resource planning, decision analysis, etc. As the data level rises, customers are focusing on cost-effective storage technologies to protect the data they are amassing on an increasing number of disk drives. For these customers, RAID provides better performance and fault tolerance than storing data on separate hard drives. RAID enables a group of disk drives to function, from the operating system perspective, like a single physical disk drive. RAID schemes, called levels, are differentiated by the method each uses to provide fault tolerance, but there is no correlation between the level numbers and the degree of fault protection. RAID 1, 1+0, 5, and Compaq RAID ADG are described in this paper because they are the levels best suited for arrays that store enterprise data. RAID 0 does not provide fault tolerance; however, it is described here in relation to RAID 1+0. This paper first describes the functions and limitations of various RAID levels in protecting data in large storage volumes. Then, it describes three of the most important factors customers should consider in their decision-making process: fault tolerance, cost effectiveness, and performance. RAID LEVELS: FUNCTIONS AND LIMITATIONS Customers needing to create large arrays with a high number of disk drives, or with high-capacity disk drives, should consider the limitations of current RAID schemes in protecting data during a single- or multiple-drive failure. This section will help customers distinguish between the different RAID levels, including RAID ADG. Table 1 on the following page summarizes these RAID technologies in regards to function, best-suited applications, and limitations. In a RAID 0 implementation, a file is split into blocks and each block is striped across all the drives in the array. For large files, reading this data in parallel from the separate drives is faster than reading the file from a single drive. However, this RAID scheme offers no fault tolerance; the entire array will fail if one drive fails. RAID 1 is a mirroring scheme that stores identical data on two sets of drives. It is used in applications that require very high availability. RAID 1 has high fault tolerance, but it has low storage efficiency because it requires twice the number of drives. RAID 1+0 is implemented as a striped array of mirrored drives. It is best suited for sites that need high performance and maximum reliability, but are willing to sacrifice storage efficiency. RAID 1+0 can withstand the failure of half the drives as long as no two drives in a mirrored pair fail; however, it sacrifices storage capacity efficiency. RAID 5 can be implemented on arrays of three or more drives. Parity information is calculated for each stripe of data and is placed on a drive other than the drive used to store the stripe of data. The parity information is striped across all drives in the array, and it occupies the equivalent capacity of one physical drive. Overall, RAID 5 provides good performance, but it can only withstand the loss of one drive without failure of the array. Compaq RAID ADG is an extension of RAID 5 for implementation on arrays of four or more drives. Data and two sets of parity information are striped across all drives in the array. The additional set of parity improves the fault tolerance of the array but results in lower write performance. The two parity sets occupy the equivalent capacity of two drives in the array. RAID ADG protects against the simultaneous failure of two drives in the array. 3