Affinity-based recovery/failover in a cluster environment

• US 7,814,065 B2
• Filed: 08/16/2005
• Issued: 10/12/2010
• Est. Priority Date: 08/16/2005
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Zoll, Michael, Roesch, Stefan, Yurek, Tolga, Chan, Wilson Wai Shun, Pruscino, Angelo
• Primary Examiner(s): Jalil; Neveen Abel
• Assistant Examiner(s): RADTKE, MARK A
• Application Number:
  US11/205,529
• Publication Number:
  US 20070043726A1
• Time in Patent Office: 1,883 Days
• Field of Search: 707/674
• US Class Current: 707/674
• CPC Class Codes: G06F 11/203 using migration
• CPC Class Codes: G06F 11/2033 switching over of hardware resources
• CPC Class Codes: G06F 11/2035 without idle spare hardware
• CPC Class Codes: G06F 11/2046 where the redundant components share persistent storage G06F11/2043 takes precedence
• CPC Class Codes: G06F 11/30 Monitoring
• CPC Class Codes: G06F 2201/825 the problem or solution involving locking
• CPC Class Codes: Y10S 707/99939 Privileged access
• CPC Class Codes: Y10S 707/99942 Manipulating data structure, e.g. compression, compaction, compilation
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

Techniques are provided for responding to the termination of a node by selecting another node, and assigning to the selected node the affinity relationships that existed between the terminated node and one or more objects. The resources that belong to the objects involved in the affinity relationships are remastered to the selected node. The selected node then performs recovery of the resources that had been opened by the terminated node and/or serves as a failover node to execute the transactions that had been executing on the terminated node.

Claims

1. A computer-implemented method for responding to termination of a terminated node, the method comprising:

using an affinity-independent assignment mechanism to cause a plurality of nodes to master all resources that belong to an object, wherein at least two resources that belong to the object are mastered by separate nodes of the plurality of nodes;

using an affinity-based assignment mechanism to (a) establish an affinity relationship between a first node of the plurality of nodes and the object and (b) cause the first node to master all resources that belong to the object;

in response to termination of the first node, determining that the affinity relationship is established between the terminated first node and object;

in response to determining that the affinity relationship is established between the terminated first node and the object, causing a particular node of the plurality of nodes to master all resources that belong to the object by transferring the affinity relationship from the first node to the particular node;

causing the particular node to perform at least one of;

recovering resources that were opened by the terminated first node; and

serving as a failover node to execute one or more transactions that had been executing on the terminated first node;

wherein the method is performed by one or more computing devices.

2. The method of claim 1 wherein causing the particular node to perform includes causing the particular node to perform both of:

recovering resources that were opened by the terminated first node; and

serving as a failover node to execute one or more transactions that had been executing on the terminated first node.

3. The method of claim 1, further comprising:

dissolving said affinity relationship between the particular node and said object in response to the particular node completing the recovery of resources that were opened by the terminated first node.

4. The method of claim 1, further comprising:

dissolving said affinity relationship between the particular node and said object in response to the particular node completing execution of the one or more transactions that had been executing on the terminated first node.

5. The method of claim 1 wherein at the time of termination of the terminated first node, the terminated first node is master of a set of certain resources for which an affinity relationship is not established between any object and said terminated first node, the method further comprising:

in response to termination of the terminated first node, remastering said set of certain resources among one or more of the plurality of nodes that remain after termination of the terminated first node.

6. The method of claim 1 wherein termination of the terminated first node is an unplanned termination caused by a failure.

7. The method of claim 1 wherein termination of the terminated first node is a planned termination.

8. The method of claim 1, further comprising:

selecting the particular node to be the recovery node for the terminated first node based on one or more characteristics of the particular node.

9. The method of claim 1, further comprising:

selecting the particular node to be the failover node for the terminated first node based on one or more characteristics of the particular node.

10. A computer-readable storage medium storing one or more sequences of instructions which, when executed by one or more processors, causes the one or more processors to perform:

using an affinity-independent assignment mechanism to cause a plurality of nodes to master all resources that belong to an object, wherein at least two resources that belong to the object are mastered by separate nodes of the plurality of nodes;

using an affinity-based assignment mechanism to (a) establish an affinity relationship between a first node of the plurality of nodes and the object and (b) cause the first node to master all resources that belong to the object;

in response to termination of the first node, determining that the affinity relationship is established between the terminated first node and the object;

in response to determining that the affinity relationship is established between the terminated first node and the object, causing a particular node of the plurality of nodes to master all resources that belong to the object by transferring the affinity relationship from the first node to the particular node;

causing the particular node to perform at least one of;

recovering resources that were opened by the terminated first node; and

serving as a failover node to execute one or more transactions that had been executing on the terminated first node.

11. The computer-readable storage medium of claim 10 wherein causing the particular node to perform includes causing the particular node to perform both of:

recovering resources that were opened by the terminated first node; and

serving as a failover node to execute one or more transactions that had been executing on the terminated first node.

12. The computer-readable storage medium of claim 10, the one or more sequences of instructions further comprising instructions for dissolving said affinity relationship between the particular node and said object in response to the particular node completing the recovery of resources that were opened by the terminated first node.

13. The computer-readable storage medium of claim 10, the one or more sequences of instructions further comprising instructions for dissolving said affinity relationship between the particular node and said object in response to the particular node completing execution of the one or more transactions that had been executing on the terminated first node.

14. The computer-readable storage medium of claim 10 wherein at the time of termination of the terminated first node, the terminated first node is master of a set of certain resources for which an affinity relationship is not established between any object and said terminated first node, the one or more sequences of instructions further comprising instructions for:

in response to termination of the terminated first node, remastering said set of certain resources among one or more of the plurality of nodes that remain after termination of the terminated first node.

15. The computer-readable storage medium of claim 10 wherein termination of the terminated first node is an unplanned termination caused by a failure.

16. The computer-readable storage medium of claim 10 wherein termination of the terminated first node is a planned termination.

17. The computer-readable storage medium of claim 10, the one or more sequences of instructions further comprising instructions for selecting the particular node to be the recovery node for the terminated first node based on one or more characteristics of the particular node.

18. The computer-readable storage medium of claim 10, the one or more sequences of instructions further comprising instructions for selecting the particular node to be the failover node for the terminated first node based on one or more characteristics of the particular node.