paritystripingofdiscarrays.docVIP

Parity Striping of Disc Arrays: Low-Cost Reliable Storage with Acceptable Throughput Jim Gray, Bob Horst, Mark Walker Tandem Computers Inc., 19333 Vallco Parkway, Cupertino, CA. 95014 Abstract An analysis of mirrored discs and of RAID5 shows that mirrors have considerably better throughput, measured as requests/second on random requests of arbitrary size (up to 1mb). Mirrors have comparable or better response time for requests of reasonable size (less than 100kb). But mirrors have a 100% storage penalty: storing the data twice. Parity striping is a data layout that stripes the parity across the discs, but does not stripe the data. Parity striping has throughput almost as good as mirrors, and has cost/gb comparable to raid5 designs -- combing the advantages of both for high-traffic disc resident data. Parity striping has additional fault containment and software benefits as well. Parity striping sacrifices the high data transfer rates of raid designs for high throughput. It is argued that response time and throughput are preferable performance metrics. Outline Introduction Why Striping and RAID Are Inappropriate for OLTP Systems Parity Striping: Cheap Reliable Storage Plus High Throughput An Analysis of Mirrors, RAID5, and Parity Stripe Performance Mirrored Discs Parity Stripe Discs RAID5 Discs Applying the Analysis to a 10+2 array The Read-Only Case The Write-Only Case Analyzing a High-Performance Disc Other Benefits of Parity Striping Summary Acknowledgments References Appendix: Spreadsheets Corresponding to the Graphs Introduction Disc arrays have traditionally been used in supercomputers to provide high transfer rates by reading or writing multiple discs in parallel [Kim]. Rather than getting 2mb/s from a single disc, applications are able to read or write N discs in parallel by striping data across the discs thereby getting a transfer rate of 2Nmb/s. The striping unit can be a bit, a byte, a sector, a page, a track, or any larger gra