System and method for modifying membership in a clustered distributed computer system and updating system configuration

• US 6,108,699 A
• Filed: 06/27/1997
• Issued: 08/22/2000
• Est. Priority Date: 06/27/1997
• Status: Expired due to Term

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Moiin, Hossein
• Primary Examiner(s): Luu, Le Hien
• Assistant Examiner(s): Prieto, Beatriz
• Application Number:
  US08/884,097
• Time in Patent Office: 1,152 Days
• Field of Search: 710/53, 710/261, 710/60, 710/260, 710/241, 710/242, 710/240, 714/1, 714/41, 714/12, 714/736, 714/14, 714/4, 709/225, 709/226, 709/201, 709/221, 709/222, 709/224, 709/252, 707/8, 370/258, 711/157, 711/152
• US Class Current: 709/221
• CPC Class Codes: G06F 11/1425 by reconfiguration of node membership

Abstract

Multiple nodes can concurrently gain membership in a cluster of nodes of a distributed computer system by broadcasting reconfiguration messages to all nodes of the distributed computer system. In response to a reconfiguration request resulting from a node petitioning to join a cluster or a node leaving the cluster, each node determines to which nodes of the distributed computer system the node is connected, i.e., which are sending reconfiguration messages which the node receives. In addition, if multiple nodes fail substantially simultaneously, each node which continues to operate does not receive a reconfiguration message from each of the failed nodes and the failed nodes are omitted from the proposed new cluster. Thus, multiple simultaneous failures are processed in a single reconfiguration. Each of the member nodes of the proposed cluster determine the membership of the proposed cluster and broadcast a reconfiguration message to all proposed member nodes and collects similar messages. If all reconfiguration messages agree, the proposed cluster is accepted. In the case in which one or more nodes leave the cluster, quorum is established in the new cluster relative to the old cluster.

Claims

1. A method for adding a new node to an existing cluster of a distributed computer system, wherein the existing cluster includes a plurality of existing nodes, said method comprising:

the new node transmitting a reconfiguration petition message to each of the plurality of existing nodes;

each of the plurality of existing nodes transmitting existing cluster configuration information to the new node and to each of the remaining of the plurality of existing nodes in response to the reconfiguration petition message;

each of the plurality of existing nodes transmitting a cluster reconfiguration message specifying a proposed new cluster to the new node and to each of the remaining of the plurality of existing nodes;

the new node transmitting an additional cluster reconfiguration message specifying an additional proposed new cluster to each of the plurality of existing nodes; and

each of the plurality of existing nodes and the new node each determining whether the proposed new clusters specified in the cluster reconfiguration messages are equivalent.

2. The method as recited in claim 1 further comprising:

each of the plurality of existing nodes and the new node each storing information which defines a new cluster in response to determining that the proposed new clusters specified in the cluster reconfiguration messages are equivalent.

3. The method as recited in claim 2 further comprising:

each of the plurality of existing nodes restoring information corresponding to said existing cluster in response to determining that the proposed new clusters specified in the cluster reconfiguration messages are not equivalent.

4. The method as recited in claim 1 further comprising:

the new node waiting a predetermined period of time for each of the plurality of existing nodes to respond to the reconfiguration petition message.

5. The method as recited in claim 1 further comprising:

each of the plurality of existing nodes and the new node each independently deriving a corresponding proposed new cluster.

6. The method as recited in claim 5, wherein each of the plurality of existing nodes derives a corresponding proposed new cluster based on the existing cluster configuration information and information identifying said new node.

7. The method as recited in claim 5, wherein the new node derives a corresponding proposed new cluster based on the existing cluster configuration information.

8. A computer readable medium comprising instructions for adding a new node to an existing cluster of a distributed computer system, wherein the existing cluster includes a plurality of existing nodes, wherein the instructions are executable by the distributed computer system to implement a method comprising:

the new node transmitting a reconfiguration petition message to each of the plurality of existing nodes;

each of the plurality of existing nodes transmitting existing cluster configuration information to the new node and to each of the remaining of the plurality of existing nodes in response to the reconfiguration petition message;

each of the plurality of existing nodes transmitting a cluster reconfiguration message specifying a proposed new cluster to the new node and to each of the remaining of the plurality of existing nodes;

the new node transmitting an additional cluster reconfiguration message specifying an additional proposed new cluster to each of the plurality of existing nodes; and

each of the plurality of existing nodes and the new node each determining whether the proposed new clusters specified in the cluster reconfiguration messages are equivalent.

9. The computer readable medium as recited in claim 8, wherein the method further comprises:

each of the plurality of existing nodes and the new node each storing information which defines a new cluster in response to determining that the proposed new clusters specified in the cluster reconfiguration messages are equivalent.

10. The computer readable medium as recited in claim 9, wherein the method further comprises:

each of the plurality of existing nodes restoring information corresponding to said existing cluster in response to determining that the proposed new clusters specified in the cluster reconfiguration messages are not equivalent.

11. The computer readable medium as recited in claim 8, wherein the method further comprises:

the new node waiting a predetermined period of time for each of the plurality of existing nodes to respond to the reconfiguration petition message.

12. The computer readable medium as recited in claim 8, wherein the method further comprises:

each of the plurality of existing nodes and the new node each independently deriving a corresponding proposed new cluster.

13. The computer readable medium as recited in claim 12, wherein each of the plurality of existing nodes derives a corresponding proposed new cluster based on the existing cluster configuration information and information identifying said new node.

14. The computer readable medium as recited in claim 12, wherein the new node derives a corresponding proposed new cluster based on the existing cluster configuration information.

15. A distributed computer system comprising:

a plurality of nodes including a plurality of processors and memory;

wherein the memory includes instructions for adding a new node to an existing cluster including a plurality of existing nodes, wherein the instructions are executable by the plurality of nodes to implement a method of;

the new node transmitting a reconfiguration petition message to each of the plurality of existing nodes;

each of the plurality of existing nodes transmitting existing cluster configuration information to the new node and to each of the remaining of the plurality of existing nodes in response to the reconfiguration petition message;

each of the plurality of existing nodes transmitting a cluster reconfiguration message specifying a proposed new cluster to the new node and to each of the remaining of the plurality of existing nodes;

the new node transmitting an additional cluster reconfiguration message specifying an additional proposed new cluster to each of the plurality of existing nodes; and

each of the plurality of existing nodes and the new node each determining whether the proposed new clusters specified in the cluster reconfiguration messages are equivalent.

16. The distributed computer system as recited in claim 15, wherein the method further comprises:

each of the plurality of existing nodes and the new node each storing information which defines a new cluster in response to determining that the proposed new clusters specified in the cluster reconfiguration messages are equivalent.

17. The distributed computer system as recited in claim 16, wherein the method further comprises:

each of the plurality of existing nodes restoring information corresponding to said existing cluster in response to determining that the proposed new clusters specified in the cluster reconfiguration messages are not equivalent.

18. The distributed computer system as recited in claim 16, wherein the method further comprises:

the new node waiting a predetermined period of time for each of the plurality of existing nodes to respond to the reconfiguration petition message.

19. The distributed computer system as recited in claim 15, wherein the method further comprises:

each of the plurality of existing nodes and the new node each independently deriving a corresponding proposed new cluster.

20. The distributed computer system as recited in claim 19, wherein each of the plurality of existing nodes derives a corresponding proposed new cluster based on the existing cluster configuration information and information identifying said new node.

21. The distributed computer system as recited in claim 19, wherein the new node derives a corresponding proposed new cluster based on the existing cluster configuration information.

22. A method for adding a new node to an existing cluster of a distributed computer system, wherein the existing cluster includes at least a first node and a second node, said method comprising:

the new node transmitting a reconfiguration petition message to the first node and the second node;

the first node transmitting existing cluster configuration information to the new node and to the second node in response to the reconfiguration petition message;

the second node transmitting existing cluster configuration information to the new node and to the first node in response to the reconfiguration petition message;

the first node transmitting a first cluster reconfiguration message specifying a first proposed new cluster to the new node and to the second node;

the second node transmitting a second cluster reconfiguration message specifying a second proposed new cluster to the new node and to the first node;

the new node transmitting a third cluster reconfiguration message specifying a third proposed new cluster to the first node and the second node; and

the first node, the second node, and the new node each determining whether the first proposed new cluster, the second proposed new cluster, and the third proposed new cluster specified in the cluster reconfiguration messages are equivalent.

23. The method as recited in claim 22 further comprising:

the first node, the second node, and the new node each storing information which defines a new cluster in response to determining that the first proposed new cluster, the second proposed new cluster, and the third proposed new cluster specified in the cluster reconfiguration messages are equivalent.

24. The method as recited in claim 23 further comprising:

the first node and the second node restoring information corresponding to said existing cluster in response to determining that the first proposed new cluster, the second proposed new cluster, and the third proposed new cluster specified in the cluster reconfiguration messages are not equivalent.

25. The method as recited in claim 22 further comprising:

the new node waiting a predetermined period of time for the first node and the second node to respond to the reconfiguration petition message.

26. The method as recited in claim 22 further comprising:

the first node deriving the first proposed new cluster independent of the second and third proposed new clusters;

the second node deriving the second proposed new cluster independent of the first and third proposed new clusters; and

the new node deriving the third proposed new cluster independent of the first and second proposed new clusters.

27. The method as recited in claim 26, wherein the first node derives the first proposed new cluster depending upon the existing cluster configuration information and information identifying said new node.

28. The method as recited in claim 26, wherein the second node derives the second proposed new cluster depending upon the existing cluster configuration information and information identifying said new node.

29. The method as recited in claim 26, wherein the new node derives the third proposed new cluster depending upon the existing cluster configuration information.

30. A computer readable medium comprising instructions for adding a new node to an existing cluster of a distributed computer system, which includes at least a first node and a second node, wherein the instructions are executable by the distributed computer system to implement a method comprising:

the new node transmitting a reconfiguration petition message to the first node and the second node;

the first node transmitting existing cluster configuration information to the new node and to the second node in response to the reconfiguration petition message;

the second node transmitting existing cluster configuration information to the new node and to the first node in response to the reconfiguration petition message;

the first node transmitting a first cluster reconfiguration message specifying a first proposed new cluster to the new node and to the second node;

the second node transmitting a second cluster reconfiguration message specifying a second proposed new cluster to the new node and to the first node;

the new node transmitting a third cluster reconfiguration message specifying a third proposed new cluster to the first node and the second node; and

the first node, the second node, and the new node each determining whether the first proposed new cluster, the second proposed new cluster, and the third proposed new cluster specified in the cluster reconfiguration messages are equivalent.

31. The computer readable medium as recited in claim 30, wherein the method further comprises:

the first node, the second node, and the new node each storing information which defines a new cluster in response to determining that the first proposed new cluster, the second proposed new cluster, and the third proposed new cluster specified in the cluster reconfiguration messages are equivalent.

32. The computer readable medium as recited in claim 31, wherein the method further comprises:

the first node and the second node restoring information corresponding to said existing cluster in response to determining that the first proposed new cluster, the second proposed new cluster, and the third proposed new cluster specified in the cluster reconfiguration messages are not equivalent.

33. The computer readable medium as recited in claim 30, wherein the method further comprises:

the new node waiting a predetermined period of time for the first node and the second node to respond to the reconfiguration petition message.

34. The computer readable medium as recited in claim 30, wherein the method further comprises:

the first node deriving the first proposed new cluster independent of the second and third proposed new clusters;

the second node deriving the second proposed new cluster independent of the first and third proposed new clusters; and

the new node deriving the third proposed new cluster independent of the first and second proposed new clusters.

35. The computer readable medium as recited in claim 34, wherein the first node derives the first proposed new cluster depending upon the existing cluster configuration information and information identifying said new node.

36. The computer readable medium as recited in claim 35, wherein the second node derives the second proposed new cluster depending upon the existing cluster configuration information and information identifying said new node.

37. The computer readable medium as recited in claim 36, wherein the new node derives the third proposed new cluster depending upon the existing cluster configuration information.

38. A distributed computer system comprising:

a plurality of nodes including a plurality of processors and memory;

wherein the plurality of nodes includes at least a first node coupled to a second node, and a new node coupled to the second node;

wherein the memory includes instructions for adding the new node to an existing cluster including said first node and second node;

wherein the instructions are executable by the plurality of nodes to implement a method of;

the new node transmitting a reconfiguration petition message to the first node and the second node;

the first node transmitting existing cluster configuration information to the new node and to the second node in response to the reconfiguration petition message;

the second node transmitting existing cluster configuration information to the new node and to the first node in response to the reconfiguration petition message;

the first node transmitting a first cluster reconfiguration message specifying a first proposed new cluster to the new node and to the second node;

the second node transmitting a second cluster reconfiguration message specifying a second proposed new cluster to the new node and to the first node;

the new node transmitting a third cluster reconfiguration message specifying a third proposed new cluster to the first node and the second node; and

the first node, the second node, and the new node each determining whether the first proposed new cluster, the second proposed new cluster, and the third proposed new cluster specified in the cluster reconfiguration messages are equivalent.

39. The distributed computer system as recited in claim 38, wherein the method further comprises:

the first node, the second node, and the new node each storing information which defines a new cluster in response to determining that the first proposed new cluster, the second proposed new cluster, and the third proposed new cluster specified in the cluster reconfiguration messages are equivalent.

40. The distributed computer system as recited in claim 39, wherein the method further comprises:

the first node and the second node restoring information corresponding to said existing cluster in response to determining that the first proposed new cluster, the second proposed new cluster, and the third proposed new cluster specified in the cluster reconfiguration messages are not equivalent.

41. The distributed computer system as recited in claim 38, wherein the method further comprises:

the new node waiting a predetermined period of time for the first node and the second node to respond to the reconfiguration petition message.

42. The distributed computer system as recited in claim 38, wherein the method further comprises:

the first node deriving the first proposed new cluster independent of the second and third proposed new clusters;

the second node deriving the second proposed new cluster independent of the first and third proposed new clusters; and

the new node deriving the third proposed new cluster independent of the first and second proposed new clusters.

43. The distributed computer system as recited in claim 42, wherein the first node derives the first proposed new cluster depending upon the existing cluster configuration information and information identifying said new node.

44. The distributed computer system as recited in claim 42, wherein the second node derives the second proposed new cluster depending upon the existing cluster configuration information and information identifying said new node.

45. The distributed computer system as recited in claim 42, wherein the new node derives the third proposed new cluster depending upon the existing cluster configuration information.