GATE CS 1999 Q2-18 — I/O Scheduling

RAID (Redundant Array of Independent Disks) is a data storage virtualization technology that combines multiple physical disk drive components into a single logical unit to achieve two primary objectives:

1. Fault-Tolerance (Reliability): Achieved by storing redundant data across multiple disks using techniques like mirroring (e.g., RAID 1) or parity (e.g., RAID 5). This ensures that the system continues to function and no data is lost even if a disk fails.
2. High Speed (Performance): Achieved by striping data across multiple disks (e.g., RAID 0). This allows read and write operations to be performed in parallel across multiple channels, significantly increasing the overall I/O throughput.

RAID does not increase data density, which is a physical characteristic of the magnetic media on individual disk platters. In fact, because of the storage overhead required for redundancy, the usable capacity of a RAID array is often less than the sum of the capacities of the individual disks.

Thus, RAID configurations are used to provide both fault-tolerance and high speed, making (D) the correct choice.