Method and apparatus for dynamically altering configurations of clustered computer systems

• US 6,769,008 B1
• Filed: 01/10/2000
• Issued: 07/27/2004
• Est. Priority Date: 01/10/2000
• Status: Expired due to Term

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Murphy, Declan J., Kumar, Krishna, Hisgen, Andrew L.
• Primary Examiner(s): Najjar, Saleh
• Application Number:
  US09/480,466
• Time in Patent Office: 1,660 Days
• Field of Search: 709/201, 709/220, 709/221, 709/222, 709/226
• US Class Current: 709/201
• CPC Class Codes: G06F 2209/505 Clust
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources
• CPC Class Codes: H04L 41/00 Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks

Abstract

Improved techniques for dynamically altering configurations of clustered computing systems are disclosed. The improved techniques can be implemented to allow alteration of an existing configuration of a clustered computing system without having to completely shutdown the clustered computing system. As a result, the clustered computing system can provide uninterrupted services while the configuration of the clustered computing system is being dynamically altered, yet also safeguard against unwanted partitions in time or space.

Claims

1. A method for altering a configuration of a clustered computing system, the clustered computing system including at least one component, said method comprising:

storing component vote information for at least one active component of the clustered computing system, wherein the component vote information includes votes assigned to components that are configured in the clustered computing system;

receiving a configuration alteration request, the configuration alteration request requesting addition or removal of one or more components to or from the existing configuration of the clustered computing system;

selecting one of the components associated with the configuration alteration request as a selected component;

obtaining a vote for the selected component; and

updating the component vote information stored for the at least one active component of the clustered computing system while the at least one active component remains active to reflect the vote for the selected component.

2. A method as recited in claim 1, wherein said updating operates to incrementally or decrementally update the component vote information with respect to an update value, and wherein said updating is performed in stages that each represent a portion of said update value.

3. A method as recited in claim 2, wherein said updating is done in increments or decrements of one.

4. A method as recited in claim 1, wherein said updating operates to prevent partitions in time or space in the clustered computing system.

5. A method as recited in claim 1,

wherein the component vote information is stored separately for each of a plurality of components of the clustered computing system, and wherein the at least one active component is one of the plurality of components, and

wherein said updating operates to serially update the component vote information stored for each of the plurality of components of the clustered computing system while the at least one active component remains active.

6. A method as recited in claim 1, wherein said method further comprises:

determining whether said updating has been successful; and

initiating an operational management process when said updating has not been successful.

7. A method as recited in claim 6,

wherein the operational management process is a method for managing operation of the clustered computing system including at least a cluster of computing nodes, and

wherein the operational management process comprises:

determining whether one of the computing nodes in the cluster has become a non-responsive node in a non-responsive sub-cluster;

determining a sub-cluster vote for a sub-cluster of one or more nodes, the sub-cluster representing a portion of the cluster that remains responsive;

obtaining a total vote for the clustered computing system;

determining whether the sub-cluster vote is at least a majority of the total vote; and

initiating shutdown of the one or more computing nodes within the sub-cluster when said determining determines that the sub-cluster vote is not at least a majority of the total vote.

8. A method as recited in claim 1, wherein said method further comprises:

determining a predetermined threshold vote; and

wherein said updating of the component vote information is performed by a series of one or more operations, each one of the one or more operations adding or subtracting the predetermined threshold vote to or from the component vote information stored for the at least one active component.

9. A method as recited in claim 8, wherein said method further comprises;

(a) determining whether one of the series of one or more operations was successful; and

(b) initiating another one of the series of one or more operations only when said determining(a) determines that one of the series of one or more operations was successful.

10. A method as recited in claim 8, wherein the predetermined threshold vote is one vote.

11. A method as recited in claim 8, wherein the selected component is a computing node or a proxy device.

12. A method as recited in claim 8,

wherein the configuration alteration request indicates that a first component and a second component are to be added to or removed from the clustered computing system,

wherein said updating operates to first update the component vote information for the first component and then secondly updates the component vote information for the second component.

13. A method as recited in claim 9,

wherein the selected component that is to be added to or removed from the clustered computing system is a proxy device, and a plurality of the components of the clustered computing system are computing nodes, and

wherein the proxy device is assigned (N−

1) votes, where N is the number of computing nodes that are connected to the proxy device.

14. A method as recited in claim 13, wherein each of the computing nodes that are connected to the proxy device is assigned one vote.

15. A method for altering a configuration of a clustered computing system including at least one component, the configuration of the clustered computing system having at least one active component, the active component having associated configuration vote information stored therein, said method comprising:

(a) receiving a configuration alteration request, the configuration alteration request requesting addition or removal of one or more components to or from the existing configuration of the clustered computing system;

(b) selecting one of the components associated with the configuration alteration request as a selected component;

(c) obtaining a vote for the selected component;

(d) updating the configuration vote information of the active component in accordance with the vote while the at least one active component remains active;

(e) determining whether said updating (d) was successful;

(f) determining whether there is another component associated with the configuration alteration request to be selected; and

(g) repeating said selecting (b) through said determining (e) for the another component when said determining (f) determines that there is another component to be selected.

16. A method as recited in claim 15, wherein said updating (d) operates to incrementally or decrementally update the component vote information in increments or decrements of one.

17. A method as recited in claim 16,

wherein the component vote information is stored separately for each of a plurality of the components of the clustered computing system, the at least one active component being one of the plurality of the components, and

wherein said updating (d) operates to serially update the component vote information associated with each of the plurality of components of the clustered computing system while the at least one active component remains active.

18. A method as recited in claim 15, wherein said method further comprises:

initiating an operational management process when said updating (d) has not been successful.

19. A clustered computing system, comprising:

a computing cluster including at least one computing node; and

a configuration manager provided for the at least one computing node, the configuration manager updating component vote information associated with at least one active component of the clustered computing system while the at least one active component remains active;

wherein the component vote information includes votes assigned to components that are configured in the clustered computing system, and

wherein the component vote information is stored for the at least one active component.

20. A clustered computing system as recited in claim 19,

wherein the configuration manager incrementally updates the component vote information when the configuration alteration request requests to add a component to the clustered computing system, and

wherein the configuration manager decrementally updates the component vote information when the configuration alteration request requests to remove a component from the clustered computing system.

21. A clustered computing system as recited in claim 20 wherein the configuration manager updates the component vote information in increments or decrements of one.

22. A clustered computing system as recited in claim 19;

wherein the clustered computing system further comprises;

an integrity protector provided on each one of the computing nodes, the integrity protector determining a vote count for a set of computing nodes in the cluster, the set of nodes representing at least a portion of the cluster, and the integrity protector determining whether the set of computing nodes should be shutdown based on the vote count.

23. A computer readable medium including computer program code for altering a configuration of a clustered computing system including at least one component, the configuration of the clustered computing system having at least one active component, the active component having associated configuration vote information stored therein, the computer readable medium comprising:

(a) computer program code for receiving a configuration alteration request, the configuration alteration request requesting addition or removal of one or more components to or from the existing configuration of the clustered computing system;

(b) computer program code for selecting one of the components associated with the configuration alteration request as a selected component;

(c) computer program code for obtaining a vote for the selected component;

(d) computer program code for updating the configuration vote information of the active component while the at least one active component remains active;

(e) computer program code for determining whether the computer program code (d) has successfully updated the configuration vote information; and

(f) computer program code for determining whether there is another component associated with the configuration alteration request to be selected.

24. A computer readable medium as recited in claim 23, wherein the computer readable medium further comprises:

(g) computer program code for repeating the computer program code (b) through the computer program code (e) when the computer program code (f) determines that there is another component to be selected.

25. A computer readable medium as recited in claim 23, wherein the computer program code (d) operates to incrementally or decrementally update the component vote information in increments or decrements of one, respectively.

26. A computer readable medium as recited in claim 23,

wherein the component vote information is stored separately for each of a plurality of the components of the clustered computing system, the at least one active component being one of the plurality of the components, and

wherein the computer program code (d) operates to serially update the component vote information associated with each of the plurality of components of the clustered computing system while the at least one active component remains active.

27. A computer readable medium as recited in claim 23, wherein the computer readable medium further comprises:

computer program code for initiating an operational management process when the computer program code (d) has not successfully updated the configuration vote information.

28. A computer readable medium including computer program code for altering configuration of a clustered computing system including at least one component, said computer readable medium comprising:

“

computer program code for storing component vote information for at least one active component of the clustered computing system, wherein the component vote information includes votes assigned to components that are configured in the clustered computing system;

computer program code for receiving a configuration alteration request, the configuration alteration request requesting addition or removal of one or more components to or from the existing configuration of the clustered computing system;

computer program code for selecting one of the components associated with the configuration alteration request as a selected component, obtaining a vote for the selected component, and

computer program code for updating the component vote information stored for the at least one active component of the clustered computing system while the at least one active component remains active to reflect the vote for the selected component.”