Method and system for dynamically reconfiguring a cluster of computer systems

• US 5,822,531 A
• Filed: 07/22/1996
• Issued: 10/13/1998
• Est. Priority Date: 07/22/1996
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Warnsman, Stuart, Gorczyca, Robert, Rashid, Aamir Arshad, Weaver, Thomas Van, Rodgers, Kevin Forress
• Primary Examiner(s): Pan, Daniel H.
• Application Number:
  US08/681,324
• Time in Patent Office: 813 Days
• Field of Search: 395/284, 395/200.62, 395/827, 395/830, 395/712, 395/182.18, 395/183.01, 395/653, 395/200.5, 384/DIG. 1, 384/DIG. 2, 379/115, 379/207, 379/34
• US Class Current: 709/221
• CPC Class Codes: G06F 2209/505 Clust
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

A method and system for dynamically reconfiguring a cluster of computer systems are disclosed. In accordance with the present invention, a first configuration file is provided at a plurality of computer systems within a cluster that specifies a current configuration of the cluster. A second configuration file is created at each of the plurality of computer systems that specifies a modified configuration of the cluster. The modified configuration is then verified. In response to the verification, the cluster is operated utilizing the modified configuration. In accordance with one embodiment, the dynamic reconfiguration of the cluster can include a reconfiguration of the cluster topology or resources.

Claims

1. A method of dynamically reconfiguring a cluster including a plurality of computer systems, said method comprising:

providing a first configuration file at each of a plurality of computer systems within a cluster, said first configuration file containing a current configuration of said cluster;

creating a second configuration file at each of said plurality of computer systems, wherein said second configuration file specifies a modified configuration of said cluster, said modified configuration including a modified cluster topology;

verifying said modified configuration, wherein verifying said modified configuration includes establishing communication with each computer system within said modified configuration and not within said current configuration; and

in response to said verification, operating said cluster utilizing said modified configuration.

2. The method of dynamically reconfiguring a cluster of claim 1, wherein said steps of creating, verifying, and operating are synchronized between said plurality of computer systems.

3. The method of dynamically reconfiguring a cluster of claim 1, and further comprising:

determining whether said creating and verifying steps were successfully completed by all of said plurality of computer systems; and

in response to a failure of any of said plurality of computer systems to complete said creating and verifying steps, aborting said dynamic reconfiguration of said cluster.

4. A method of dynamically reconfiguring a cluster including a plurality of computer systems, said method comprising:

providing a first configuration file at each of a plurality of computer systems within a cluster, said first configuration file containing a current configuration of said cluster;

creating a second configuration file at each of said plurality of computer systems, wherein said second configuration file at each of said plurality of computer systems specifies a modified configuration including a diverse set of cluster resources than specified by said first configuration file;

verifying said modified configuration; and

in response to said verification, operating said cluster utilizing said modified configuration.

5. The method of dynamically reconfiguring a cluster of claim 4, wherein said step of verifying said modified configuration comprises verifying ownership by said plurality of computer systems of selected resources accessible under said modified configuration.

6. The method of dynamically reconfiguring a cluster of claim 4, wherein said steps of creating, verifying, and operating are synchronized between said plurality of computer systems.

7. The method of dynamically reconfiguring a cluster of claim 4, and further comprising:

determining whether said creating and verifying steps were successfully completed by all of said plurality of computer systems; and

in response to a failure of any of said plurality of computer systems to complete said creating and verifying steps, aborting said dynamic reconfiguration of said cluster.

8. A cluster data processing system, comprising:

a plurality of computer systems, each of said plurality of computer systems including a memory;

a communications link connecting each of said plurality of computer systems with at least one other of said plurality of computer systems for communication;

a first configuration file stored within memory at each of said plurality of computer systems, said first configuration file containing a current configuration of said cluster; and

a cluster manager stored within said memory of at least one of said plurality of computer systems and executable by at least one of said plurality of computer systems, wherein said cluster manager creates a second configuration file at each of said plurality of computer systems, said second configuration file specifying a modified configuration of said cluster, and wherein said cluster manager verifies said modified configuration of said cluster, and in response to said verification, operates said cluster utilizing said modified configuration.

9. The cluster a data processing system of claim 8 wherein said second configuration file created at each of said plurality of computer systems by said cluster manager specifies a modified cluster topology.

10. The cluster data processing system of claim 9, wherein said cluster manager verifies said modified configuration by establishing communication with each computer system that is within said modified configuration and not within said current configuration.

11. The cluster data processing system of claim 8, wherein said second configuration file created at each of said plurality of computer systems by said cluster manager specifes a diverse set of cluster resources than said first configuration file.

12. The cluster data processing system of claim 11, wherein said cluster manager verifies said configuration by verifying ownership by said plurity of computer systems of selected resources accessable under said modified configuration.

13. The cluster data processing system of claim 8, wherein said cluster manager synchronizes creation of said second configuration file, verification of said modified configuration, and operation of said cluster utilizing said modified configuration between said plurity of computer systems.

14. The cluster data processing system of claim 8, wherein said cluster manager aborts said dynamic reconfiguration of said cluster in response to a failure of any of said plurity of computer systems to successfully complete creation of said second configuration file and verification of said modified configuration.

15. A program product for dynamically reconfiguring a cluster including a plurality of computer systems, said program product comprising:

a cluster manager executable by at least one of a plurality of computer systems within a cluster, each of said plurality of computer systems within said cluster having a first configuration file containing a current configuration of said cluster, wherein said cluster manager creates a second configuration file at each of said plurality of computer systems, said second configuration file specifying a modified configuration including a modified cluster topology, and wherein said cluster manager verifies said modified configuration of said cluster, and in response to said verification, operates said cluster utilizing said modified configuration; and

an information bearing media bearing said cluster manager.

16. The program product of claim 15, wherein said cluster manager verifies said modified configuration by establishing communication with each computer system that is within said modified configuration and not within said current configuration.

17. The program product of claim 15, wherein said cluster manager synchronizes creation of said second configuration file, verification of said modified configuration, and operation of said cluster utilizing said modified configuration between said plurality of computer systems.

18. The program product of claim 15, wherein said cluster manager aborts dynamic reconfiguration of said cluster in response to a failure of any of said plurality of computer systems to successfully complete creation of said second configuration file and verification of said modified configuration.

19. A program product for dynamically reconfiguring a cluster including a plurality of computer systems, said program product comprising:

a cluster manager executable by at least one of a plurality of computer systems within a cluster, each of said plurality of computer systems within said cluster having a first configuration file containing a current configuration of said cluster, wherein said cluster manager creates a second configuration file at each of said plurality of computer systems, said second configuration file specifying a modified configuration including a diverse set of cluster resources than specified by said first configuration file, and wherein said cluster manager verifies said modified configuration of said cluster, and in response to said verification, operates said cluster utilizing said modified configuration; and

an information bearing media bearing said cluster manager.

20. The program product of claim 19, wherein said cluster manager verifies said modified configuration by verifying ownership by said plurality of computer systems of selected resources accessible under said modified configuration.

21. The program product of claim 19, wherein said cluster manager synchronizes creation of said second configuration file, verification of said modified configuration, and operation of said cluster utilizing said modified configuration between said plurality of computer systems.

22. The program product of claim 19, wherein said cluster manager aborts dynamic reconfiguration of said cluster in response to a failure of any of said plurality of computer systems to successfully complete creation of said second configuration file and verification of said modified configuration.