a) How is each component of the disk-access time effected by the disk's RPM (revolutions per minute)?
b) How is each component of the disk-access time effected by the sector size?
2. Suppose we have a 5 disk RAID 5 (block-level distributed parity) array.
a) If Disk 0 crashes, reconstruct the first three bits of block 0. Assume even parity is being used.
b) Assume that Disk 0 has crashed. If you write a new block of data starting with bits 0 1 1 ... to Block 3, then can Blocks 5 and 6 be read at the same time as the write for Block 3 is occurring? (Justify your answer)
3. Suppose we have an 5 disk RAID array with each disk having a 100 MB/sec data transfer rate. Complete the following table assuming none of the disks are faulty.
RAID Level | Maximum number of concurrent, independent READs | Maximum number of concurrent, independent WRITEs | Data Transfer Rate for a single large READ |
RAID 1 (Mirroring with large strips) |   |   |   |
RAID 3 (bit-interleaved parity) |   |   |   |
RAID 5 (block-level distributed parity) |   |   |