Data fault tolerance software apparatus and method

• US 6,427,212 B1
• Filed: 11/13/1998
• Issued: 07/30/2002
• Est. Priority Date: 11/13/1998
• Status: Expired due to Term

• Current Assignee: PMC-Sierra Incorporated (Microchip Technology Incorporated)
• Original Assignee: Tricord Systems, Inc. (Steel Partners Holdings LP)
• Inventors: Frey, Alexander H. Jr.
• Primary Examiner(s): Wong, Peter
• Assistant Examiner(s): VO, TIM T
• Application Number:
  US09/191,943
• Time in Patent Office: 1,355 Days
• Field of Search: 714/6, 711/114, 711/170, 711/112, 711/162, 711/7
• US Class Current: 714/6.24
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 2211/1028 Distributed, i.e. distributed RAID systems with parity

Abstract

The invention discloses apparatus and process in which data files are distributed across a large scale data processing system to enable protection from the loss of data due to the failure of one or more fault domains. Specifically, the invention provides significant advances in data base management by distributing data across N fault domains using one or more of a multitude of deterministic functions to protect failure.

Claims

1. A data fault tolerance software system comprising:

means for organizing data into a data file having at least one data block, wherein each data block is up to a predetermined length;

means for organizing protection information into a protection file having at least one protection block, wherein each of the at least one data block is associated with one of the at least one protection block, the data file and protection file being different file objects;

means for identifying each data file and each protection file; and

means for changing a protection block In the protection file without requiring a rewrite of a data block.

2. The software system of claim 1 wherein said predetermined length is determined by a user and is independent of the system'"'"'s configuration.

3. The software system of claim 1 wherein said means for organizing said data file includes a starter node means that starts the data file with a next data block being placed in a next node and placement continues in this manner.

4. The software system of claim 1 wherein said at least one protection block includes a parameter which is independent of the position at which any particular data block is stored.

5. The software system of claim 2 wherein said predetermined length is independent of the existence of an algorithm for parity protection.

6. The software system of claim 3 wherein said placement is implemented to structure consecutive data blocks on consecutively numbered nodes until a last node of a pre-specified span is reached.

7. The software system of claim 4 wherein a single one of said protection blocks is implemented to protect multiple data blocks.

8. The software system of claim 6 wherein said span designates the number of nodes over which the data file will be spread.

9. The software system of claim 7 wherein said protection blocks utilize a protection scheme comprising an exclusive-OR of data.

10. A data fault tolerance software system implemented in a microprocessor, wherein the software system utilizes a process to provide data fault tolerance, the microprocessor-implemented process comprising:

a data spreading process for spreading data into a data file having at least one data block;

a protection spreading process for spreading redundancy information into a redundancy file having at least one redundancy block, said at least one redundancy block associated with at least one data block; and

a data redundancy changing process to change a redundancy block without rewriting a data block.

11. A method, comprising:

storing one or more data blocks in a data file;

storing one or more parity protection blocks in a parity file, each of said one or more parity protection blocks including parity information in accordance with a first protection scheme and based on the contents of at least one of the one or more data blocks; and

changing the first protection scheme to a second protection scheme by updating at least one of the one or more parity protection blocks without rewriting the data blocks.

12. The method of claim 11, wherein the act of storing the one or more data blocks comprises storing the one or more data blocks in accordance with an exclusive-OR scheme.

13. The method of claim 11, wherein the act of changing the first protection scheme comprises changing the first protection scheme to a mirroring scheme.

14. The method of claim 11, wherein the act of storing one or more protection blocks comprises storing parity bits in the protection file that are generated based on the one or more data blocks.

15. The method of claim 12, wherein the act of updating comprises substituting a parity protection block in accordance with the second protection scheme for the parity block in accordance with the first protection scheme.

16. The method of claim 14, further comprising changing the parity bits in a protection block in the protection file without rewriting the one or more data blocks.