Apparatus and method for maintaining consistency of shared data resources in a cluster environment

• US 6,748,381 B1
• Filed: 03/31/1999
• Issued: 06/08/2004
• Est. Priority Date: 03/31/1999
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Shaheen, Amal Ahmed, Chao, Ching-Yun, Hough, Roger Eldred
• Primary Examiner(s): Alam, Shahid
• Assistant Examiner(s): FLEURANTIN, JEAN B
• Application Number:
  US09/282,908
• Time in Patent Office: 1,896 Days
• Field of Search: 707/1-2, 707/9-10, 707/200-205, 709/221, 709/213, 709/223, 709/917, 709/244, 714/10-11, 714/13
• US Class Current: 1/1
• CPC Class Codes: G06F 11/182 based on mutual exchange of the output between redundant processing components
• CPC Class Codes: G06F 16/2365 Ensuring data consistency and integrity
• CPC Class Codes: H04L 67/104 Peer-to-peer [P2P] networks
• CPC Class Codes: H04L 67/1046 Joining mechanisms
• CPC Class Codes: H04L 67/1095 Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes
• CPC Class Codes: Y10S 707/99952 Coherency, e.g. same view to multiple users

Abstract

A method and apparatus for providing a recent set of replicas for a cluster data resource within a cluster having a plurality of nodes; each of the nodes having a group services client with membership and voting services. The method broadcasts a data resource open request to the nodes of the cluster, determines a recent replica of the cluster data resource among the nodes, and distributes the recent replica to the nodes of the cluster. The apparatus is for providing a recent set of replicas for a cluster data resource. The apparatus has a cluster having a plurality of nodes in a peer relationship; each node has an electronic memory for storing a local replica of the cluster data resource. A group services client, which is executable by each node of the cluster, has cluster broadcasting and cluster voting capability. A database conflict resolution protocol (“DCRP”), which is executable by each node of the cluster, interacts with the group services clients such that the DCRP broadcasts to the nodes a data resource modification request having a data resource identifier and a timestamp. The DCRP determines a recent replica of the cluster data resource among the nodes with respect to the timestamp of the broadcast data resource modification request relative to a local timestamp associated with the data resource identifier, and distributes the recent replica of the cluster data resource to each node of the plurality of nodes.

Claims

1. A method for providing a recent set of replicas for a cluster data resource within a cluster having a plurality of nodes each having a group services client with membership and voting services to the cluster, the method comprising the steps of:

broadcasting to the plurality of nodes a data resource open request having a data resource identifier and a timestamp;

determining a recent replica of the cluster data resource among the plurality of nodes with respect to the timestamp of the broadcast data resource open request relative to a local timestamp of a node of the plurality of nodes associated with the data resource identifier; and

distributing the recent replica of the cluster data resource to each node of the plurality of nodes.

2. The method of claim 1 wherein the step of determining further comprises:

determining relative to a local timestamp and a local checksum.

3. The method of claim 1 wherein the data resource open request is issued by the group services client of a node of the plurality of nodes in response to an application client executing on the node.

4. The method of claim 1 wherein the data resource open request is issued in a series of time intervals.

5. The method of claim 1 wherein the data resource open request is issued when an additional node joins the cluster.

6. The method of claim 1 wherein the step of determining further comprises the steps of:

comparing the timestamp of the broadcast data resource open request relative to the local timestamp of a node of the plurality of nodes associated with the data resource identifier; and

voting a comparison result to the plurality of nodes.

7. The method of claim 6, further comprising the step of:

repeating the steps of comparing and voting until a favorable consensus is reached.

8. The method of claim 6 wherein the favorable consensus is an APPROVE vote by each node of the plurality of nodes.

9. The method of claim 6 wherein the favorable consensus is an APPROVE vote by all but one node of the plurality of nodes.

10. An apparatus for providing a recent set of replicas for a cluster data resource within a cluster having a plurality of nodes each having a group services client with membership and voting services to the cluster, the method comprising the steps of:

means for broadcasting to said plurality of nodes a data resource open request having a data resource identifier and a timestamp;

means for determining a recent replica of said cluster data resource among said plurality of nodes with respect to said timestamp of said broadcast data resource open request relative to a local timestamp of a node of the plurality of nodes associated with the data resource identifier; and

means for distributing the recent replica of the cluster data resource to each node of the plurality of nodes.

11. The apparatus of claim 10 wherein the data resource open request is issued by the group services client of a node of the plurality of nodes in response to an application client executing on said node.

12. The apparatus of claim 10 wherein the data resource open request is issued according to a series of time intervals.

13. The apparatus of claim 10 wherein the data resource open request is issued when an additional node joins the cluster.

14. The apparatus of claim 10 wherein the determining means further comprise:

means for comparing the timestamp of the broadcast data resource open request relative to the local timestamp of a node of the plurality of nodes associated with the data resource identifier; and

means for voting a comparison result to the plurality of nodes.

15. An apparatus for providing a recent set of replicas for a cluster data resource, the apparatus comprising:

a cluster having a plurality of nodes in a peer relationship, each node of said plurality of nodes having an electronic memory for storing a local replica of the cluster data resource,

a group services client executable by each node of said plurality of nodes, said group services client having a cluster broadcasting and a cluster voting capability; and

a database conflict resolution protocol (“

DCRP”

) executable by each node of said plurality of nodes, said DCRP interacting with said group services clients such that said DCRP broadcasts to said plurality of nodes a data resource open request having a data resource identifier and a timestamp, determines a recent replica of said cluster data resource among said plurality of nodes with respect to said timestamp of said broadcast data resource open request relative to a local timestamp of a node of said plurality of nodes associated with said data resource identifier, and distributes said recent replica of said cluster data resource to each node of said plurality of nodes.

16. A method for establishing a consistent set of replicas of a database within a computer cluster using cluster membership and cluster voting services, comprising the steps of:

broadcasting an open request to other nodes to locate their corresponding replicas;

voting to select a most recently modified replica of the database;

determining through the voting of a source of the selected replica to be used by all participating nodes; and

broadcasting the selected replica to participating nodes with differing or non-existent replicas;

wherein a most recently modified replica is determined using a timestamp;

wherein a checksum or cyclic redundancy value is used to further ensure consistency using the checksum or cyclic redundancy value as a low-order portion of the timestamp, and updating the checksum or cyclic redundancy value on each update of the database.

17. The method as recited in claim 16, wherein the checksum or cyclic redundancy value is computed using all data values within the database replica, but not including the timestamp of the last modification.