Method and system for improved efficiency of parity calculation in RAID system

Patent 6282671 Issued on August 28, 2001. Estimated Expiration Date:

November 10, 2018. Estimated Expiration Date is calculated based on simple USPTO term provisions. It does not account for terminal disclaimers, term adjustments, failure to pay maintenance fees, or other factors which might affect the term of a patent.

Abstract Claims Description Full Text

Patent References

3905023

Optimized stripe detection for redundant arrays of disk drives
Patent #: 5315602
Issued on: 05/24/1994
Inventor: Noya, et al.

System and method for selective write-back caching within a disk array subsystem
Patent #: 5572660
Issued on: 11/05/1996
Inventor: Jones

XOR controller for a storage subsystem
Patent #: 5594862
Issued on: 01/14/1997
Inventor: Winkler, et al.

Information handling system using default status conditions for transfer of data blocks Patent #: 5794069
Issued on: 08/11/1998
Inventor: Chisholm, et al.

Inventors

Application

No. 189689 filed on 11/10/1998

US Classes:

714/6, Redundant stored data accessed (e.g., duplicated data, error correction coded data, or other parity-type data)711/114Arrayed (e.g., RAIDs)

Examiners

Primary: Ray, Gopal C.

Attorney, Agent or Firm

Bracewell & Patterson, LLP

International Classes

G06F 011/10
H03M 013/00

Abstract

The method and system for improved efficiency and parity calculation in RAID data storage system. RAID data storage systems permit recovery of data in the event of a drive failure by storing a parity value for each stripe of data on a separate drive. Each time data is written to a RAID data storage the parity value for the associated data stripe must be updated. Parity may be calculated utilizing two different algorithms. New parity may be calculated utilizing the old data value, the new data value and the old parity value, or calculated utilizing the new data value and all remaining old data within the associated data stripe. The number of input/output operations required to calculate parity utilizing each algorithm is predetermined and the method requiring the least number of input/output operations is selected, unless data sequentiality is detected. If sequential data is present the parity value is calculated utilizing the new data and all old data in the associated data stripe, even if a greater number of input/output operations are initially required, as further sequential input/output operations will be more efficient.