Technique for project partitioning in a cluster of servers

• US 7,516,181 B1
• Filed: 02/08/2005
• Issued: 04/07/2009
• Est. Priority Date: 02/08/2005
• Status: Active Grant

• Current Assignee: Microstrategy Incorporated
• Original Assignee: Microstrategy Incorporated
• Inventors: Li, Benjamin Z., Du, Yi, Bui, Quoc Cuong, Liu, Ningning, Polana, Ramprasad
• Primary Examiner(s): WANG, LIANG CHE A
• Application Number:
  US11/052,469
• Time in Patent Office: 1,519 Days
• Field of Search: 709/205, 709219-229
• US Class Current: 709/205
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/1014 based on the content of a request
• CPC Class Codes: H04L 67/1034 Reaction to server failures by a load balancer

Abstract

A technique for project partitioning in a cluster of servers is disclosed. In one particular exemplary embodiment, the technique may be realized as a method for partitioning a plurality of projects among a cluster of servers. The method may comprise determining a status of project partition among the cluster of servers. The method may also comprise generating a partition plan for re-partitioning the plurality of projects among the cluster of servers based at least in part on the status and a partition policy, wherein the partition policy comprises one or more rules selected from a group consisting of a user specified project-server association rule, a weighted balancing rule, and a project redundancy rule. The method may further comprise communicating the partition plan to each of the cluster of servers. The method may additionally comprise causing the plurality of projects to be re-partitioned among the cluster of servers based on the partition plan.

Claims

1. A method for partitioning a plurality of projects among a cluster of servers, the method comprising:

determining a status of project partition among the cluster of servers, wherein the cluster of servers hosts a common set of projects and each server of the cluster of servers hosts a subset of the common set of projects;

generating a partition plan for re-partitioning the plurality of projects among the cluster of servers based at least in part on the status and a partition policy, wherein the partition policy comprises one or more rules selected from a group consisting of;

a user specified project-server association rule,

a weighted balancing rule, and

a project redundancy rule;

communicating the partition plan to each of the cluster of servers; and

causing the plurality of projects to be re-partitioned among the cluster of servers based on the partition plan.

2. The method according to claim 1 being triggered by one or more events selected from a group consisting of:

a server leaving the cluster;

an additional server joining the cluster;

failure of at least one server in the cluster;

restart of at least one server in the cluster; and

a user trigger via an application program interface.

3. The method according to claim 1 further comprising:

re-partitioning, after a predetermined delay, the plurality of projects among the cluster of servers if a server running one or more projects becomes unavailable and no duplicate for the one or more projects exists on any other server.

4. The method according to claim 1 further comprising:

selecting, based on an election algorithm, a surviving server of the cluster of servers to generate the partition plan when a server of the cluster of servers becomes unavailable.

5. The method according to claim 1, wherein the status of project partition among the cluster of servers is recorded in a partition matrix.

6. The method according to claim 5, wherein the partition plan comprises a transformed matrix based on the partition matrix.

7. The method according to claim 6 further comprising:

communicating the transformed matrix to each server in the cluster.

8. The method according to claim 1 further comprising:

specifying, through a user interface, an initial project partition among the cluster of servers.

9. The method according to claim 1, wherein the weighted balancing rule causes each project to be distributed to at least a predetermined number of servers in the cluster.

10. The method according to claim 1, wherein the weighted balancing rule causes each project to be distributed to at least a predetermined percentage of servers in the cluster.

11. The method according to claim 1, wherein each server immediately responds to the partition plan upon receiving the partition plan.

12. The method according to claim 1, wherein each server responds to the partition plan after a predetermined delay.

13. The method according to claim 1, wherein the cluster of servers comprise one or more intelligence servers.

14. A system for partitioning a plurality of projects among a cluster of servers, the system comprising:

a plurality of project partition modules, each residing in one of the cluster of servers, wherein each project partition module is adapted to;

determine a status of project partition among the cluster of servers, wherein the cluster of servers hosts a common set of projects and each server of the cluster of servers hosts a subset of the plurality of projects;

generate a partition plan for re-partitioning the plurality of projects among the cluster of servers based at least in part on the status and a partition policy, wherein the partition policy comprises one or more rules selected from a group consisting of;

a user specified project-server association rule,

a weighted balancing rule, and

a project redundancy rule;

communicate the partition plan to other project partition modules in the cluster; and

cause the plurality of projects to be re-partitioned among the cluster of servers based on the partition plan.

15. An article of manufacture for partitioning a plurality of projects among a cluster of servers, the article of manufacture comprising:

at least one processor readable carrier; and

instructions carried on the at least one carrier;

wherein the instructions are configured to be readable from the at least one carrier by at least one processor and thereby cause the at least one processor to operate so as to;

determine a status of project partition among the cluster of servers, wherein the cluster of servers hosts a common set of projects and each server of the cluster of servers hosts a subset of the plurality of projects;

generate a partition plan for re-partitioning the plurality of projects among the cluster of servers based at least in part on the status and a partition policy, wherein the partition policy comprises one or more rules selected from a group consisting of;

a user specified project-server association rule,

a weighted balancing rule, and

a project redundancy rule;

communicate the partition plan to each of the cluster of servers; and

cause the plurality of projects to be re-partitioned among the cluster of servers based on the partition plan.