Method, system and program products for recovering from failures within a shared nothing distributed computing environment

• US 6,873,987 B1
• Filed: 05/31/2000
• Issued: 03/29/2005
• Est. Priority Date: 05/31/2000
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Novaes, Marcos N., Laib, Gregory D.
• Primary Examiner(s): Rones, Charles
• Assistant Examiner(s): Mahmoudi, Hassan
• Application Number:
  US09/583,784
• Time in Patent Office: 1,763 Days
• Field of Search: 371/7, 395/575, 395/670, 395/182, 707/8, 707/10, 707/202, 707/205, 711/150, 711/153, 712/13, 714/4, 709/104, 709/201
• US Class Current: 707/611
• CPC Class Codes: G06F 11/1482 by means of middleware or OS functionality
• CPC Class Codes: Y10S 707/966 Distributed
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

The management and use of replicated distributed transactions is facilitated. A distributed synchronous transaction system protocol is provided to manage the replication of distributed transactions for client application instances. The distributed synchronous transaction system allows transactions to be replicated without having the client application instances be aware of other instances to receive the transaction. Further, if a failure occurs during processing of a distributed replicated transaction, the distributed synchronous transaction system manages the recovery of the failure.

Claims

1. A method of recovery from failures within a shared nothing distributed computing environment, said method comprising:

detecting multiple overlapping failures within said shared nothing distributed computing environment; and

automatically recovering from said multiple overlapping failures, wherein one or more transactions affected by one or more failures of said multiple overlapping failures are automatically executed to completion without rolling back said one or more transactions and without requiring a reposting of said one or more transactions.

2. The method of claim 1, wherein the shared nothing distributed computing environment comprises a processing group with a plurality of members, and wherein the detecting comprises detecting multiple overlapping failures of at least some of the plurality of members.

3. The method of claim 2, wherein the recovering comprises synchronizing messages regarding the one or more transactions among surviving members of the processing group.

4. The method of claim 3, wherein the recovering further comprises committing the one or more transactions.

5. The method of claim 2, wherein at least one member of the processing group survives the failures, and wherein the recovering comprises electing a coordinator from among the at least one surviving member.

6. The method of claim 5, wherein the recovering further comprises receiving by the coordinator a list of the one or more transactions from the other surviving members.

7. The method of claim 6, wherein the recovering further comprises receiving by the coordinator any commit protocol messages for the one or more transactions the coordinator does not already have.

8. The method of claim 7, wherein the coordinator initiates the commit protocol for the one or more transactions.

9. The method of claim 1, wherein the shared nothing distributed computing environment comprises a distributed synchronous transaction system, and wherein the method comprises a failure recovery method for the distributed synchronous transaction system.

10. A system of recovery from failures within a shared nothing distributed computing environment, said system comprising:

means for detecting multiple overlapping failures within said shared nothing distributed computing environment; and

means for automatically recovering from said multiple overlapping failures, wherein one or more transactions affected by one or more failures of said multiple overlapping failures are automatically executed to completion without rolling back said one or more transactions and without requiring a reposting of said one or more transactions.

11. The system of claim 10, wherein the shared nothing distributed computing environment comprises a processing group with a plurality of members, and wherein the means for detecting comprises means for detecting multiple overlapping failures of at least some of the plurality of members.

12. The system of claim 11, wherein the means for recovering comprises means for synchronizing messages regarding the one or more transactions among surviving members of the processing group.

13. The system of claim 12, wherein the means for recovering further comprises means for committing the one or more transactions.

14. The system of claim 11, wherein at least one member of the processing group survives failures, and wherein the means for recovering comprises means for electing a coordinator from among the at least one surviving member.

15. The system of claim 14, wherein the means for recovering further comprises means for receiving by the coordinator an indication of the one or more transactions from the other surviving members.

16. The system of claim 15, wherein the means for recovering further comprises means for receiving by the coordinator any commit protocol messages for the one or more transactions the coordinator does not already have.

17. The system of claim 16, wherein the means for recovering further comprises means for the coordinator to initiate the commit protocol for the one or more transactions.

18. At least one program storage device readable by a machine, tangibly embodying at least one program of instructions executable by the machine to perform a method of recovery from failures within a shared nothing distributed computing environment, said method comprising:

detecting multiple overlapping failures within said shared nothing distributed computing environment; and

automatically recovering from said multiple overlapping failures, wherein one or more transactions affected by one or more failures of said multiple overlapping failures are automatically executed to completion without rolling back said one or more transactions and without requiring a reposting of said one or more transactions.

19. The at least one program storage device of claim 18, wherein the shared nothing distributed computing environment comprises a processing group with a plurality of members, and wherein the detecting comprises detecting multiple overlapping failures of at least some of the plurality of members.

20. The at least one program storage device of claim 19, wherein the recovering comprises synchronizing messages regarding the one or more transactions among surviving members of the processing group.

21. The at least one program storage device of claim 20, wherein the recovering further comprises committing the one or more transactions.

22. The at least one program storage device of claim 19, wherein at least one member of the processing group survives failures, and wherein the recovering comprises electing a coordinator from among the at least one surviving member.

23. The at least one program storage device of claim 22, wherein the recovering further comprises receiving by the coordinator a list of the one or more transactions from the other surviving members.

24. The at least one program storage device of claim 23, wherein the recovering further comprises receiving by the coordinator any commit protocol messages for the one or more transactions the coordinator does not already have.

25. The at least one program storage device of claim 24, wherein the coordinator initiates the commit protocol for the one or more transactions.