DEPLOYMENT OF PRE-SCHEDULED TASKS IN CLUSTERS

• US 20130067493A1
• Filed: 09/09/2011
• Published: 03/14/2013
• Est. Priority Date: 09/09/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Dion, David A., Koppolu, Lokesh S., Synytskyy, Mykyta, Perriman, Symon, Sun, Zongwei, Yanchenko, Maksym
• Granted Patent: US 8,875,157 B2
• Time in Patent Office: Days
• US Class Current: 719/318
• CPC Class Codes: G06F 9/4843 by program, e.g. task dispatcher, supervisor, operating system

Abstract

Managing a triggerable task in a cluster environment. A method includes accessing a definition a task, wherein the definition includes a triggering event, and an action to be performed when the triggering event occurs. The method further includes, accessing a cluster policy defining how triggering events are interpreted on a node task basis for a plurality of nodes in a cluster. Based on the cluster policy, the method includes determining that the triggering event has occurred. As a result of the determination that the triggering event has occurred, the method further includes attempting to perform the action.

Claims

1. A method of managing a triggerable task in a cluster environment:

accessing a definition of a task, wherein the definition includes a triggering event, and one or more actions to be performed when the triggering event occurs;

accessing a cluster policy defining how triggering events are interpreted on a node/task basis for a plurality of nodes in a cluster;

based on the cluster policy, determining that the triggering event has occurred; and

as a result attempting to perform the one or more actions.

2. The method of claim 1, wherein the triggering event comprises at least one of cluster service start-up or shut-down.

3. The method of claim 1, wherein the triggering event comprises at least one of a network connection failure, SAN connectivity failure, storage corruption, low storage capacity, specific time of day, or expiration of pre-determined period of timec.

4. The method of claim 1, wherein the cluster policy defines that one or more of the actions should be run on at least one of only one node, every node, or nodes running a particular application.

5. The method of claim 1, wherein the cluster policy defines that one or more of the actions should be run at least once or run at most once.

6. The method of claim 1, further comprising storing task state if the one or more actions are not completely performed.

7. The method of claim 1, further comprising attempting to perform the one or more actions until the one or more actions are performed successfully.

8. The method of claim 1, further comprising only attempting to perform one or more of the one or more actions at most once no matter whether the actions are successful or not.

9. The method of claim 1, further comprising attempting to perform one or more of the one or more actions on a different node if actions fail on a first node.

10. The method of claim 1, further comprising attempting to perform one or more of the one or more actions and placing an application tied to the task on a different node if actions fail on a first node.

11. The method of claim 1, wherein the cluster policy including cluster task configuration and/or status data is added to one or more nodes added to the cluster even though the task was defined before the one or more node were added to the cluster.

12. The method of claim 1, wherein the cluster policy including cluster task configuration and/or status data is removed from one or more nodes removed from the cluster.

13. A computer readable medium comprising computer executable instructions that when executed by one or more processor causes the one or more processors to perform the following acts for managing a triggerable task in a cluster environment:

accessing a task definition, wherein the definition defines a trigger and one or more actions to be performed when the trigger occurs;

accessing a cluster policy defining how triggers are handled for different nodes in the cluster;

based on the cluster policy, determining that the trigger has occurred; and

as a result attempting to perform the one or more actions.

14. The computer readable medium of claim 13, wherein the cluster policy defines that one or more of the actions should be run on at least one of only one node, every node, or nodes running a particular application.

15. The computer readable medium of claim 13, wherein the cluster policy defines that one or more of the actions should be run at least once or run at most once.

16. The computer readable medium of claim 13, further comprising attempting to perform the one or more actions until the one or more actions are performed successfully.

17. The computer readable medium of claim 13, further comprising only attempting to perform one or more of the one or more actions once no matter whether the actions are successful or not.

18. The computer readable medium of claim 13, further comprising attempting to perform one or more of the one or more actions on a different node if actions fail on a first node.

19. The computer readable medium of claim 13, further comprising attempting to perform one or more of the one or more actions and placing an application tied to the task on a different node if actions fail on a first node.

20. A distributed computing system, the computing system comprising:

a plurality of computers wherein the computer are organized into a computing cluster as nodes of the computing cluster, wherein each of the nodes comprises;

a cluster service, the cluster service having triggerable cluster tasks created on the cluster service;

a cluster database, wherein the cluster database on each of the nodes is synchronized with cluster databases on other nodes in the cluster, and wherein the cluster database stores triggerable cluster task properties; and

a task scheduler service, wherein the task scheduler service interacts with the cluster service to creates cluster triggerable tasks locally on a local node.