High-availability computer cluster with failover support based on a resource map

• US 8,738,961 B2
• Filed: 08/17/2010
• Issued: 05/27/2014
• Est. Priority Date: 08/17/2010
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Jain, Reshu, Sarkar, Prasenjit, Seaman, Mark James
• Primary Examiner(s): Leibovich, Yair
• Application Number:
  US12/858,436
• Publication Number:
  US 20120047394A1
• Time in Patent Office: 1,379 Days
• Field of Search: 714/4.11
• US Class Current: 714/4.11
• CPC Class Codes: G06F 11/2028 eliminating a faulty processor or activating a spare

Abstract

Embodiments of the invention relate to handling failures in a cluster of computer resources. The resources are represented as nodes in a dependency graph in which some nodes are articulation points and the removal of any articulation point due to a resource failure results in a disconnected graph. The embodiments perform a failover when a resource corresponding to an articulation point fails. The failover is to a local resource if the failed resource does not affect all local resources. The failover is to a remote resource if no local resource can meet all resource requirements of the failed resource, and to a remote resource running in a degraded mode if the remote resource cannot meet all of the requirements.

Claims

1. A method comprising:

representing a cluster of computer resources as nodes in a dependency graph, the nodes including a plurality of articulation points, wherein removal of an articulation point due to a resource failure results in a disconnected dependency graph;

if a failed resource corresponds to an articulation point, performing a failover for a resource group that the failed resource is a member, wherein all of the nodes in a cluster are aware of status of all resources on all other nodes, and

if a failover remote resource only satisfies a portion of all resource requirements, then failover still occurs and the failover remote resource runs in a degraded mode,

wherein the resources comprise local resources and the failover is to a local resource if the failed resource does not affect all of the local resources, and upon a resource failing, analyzing the dependency graph to determine if an operational state of a primary node is reachable.

2. The method of claim 1, wherein the resources comprise remote resources, each resource is associated with a plurality of resource requirements, and the failover is to a remote resource if the remote resource can meet the resource requirements of the failed resource, wherein all of the nodes in a cluster are constantly aware of status of all resources on all other nodes.

3. The method of claim 2, wherein, performing a failover comprises performing failover for resources of an entire resource group that the failed resource is a member.

4. The method of claim 2, wherein the remote resource is part of a remote computer cluster, wherein failover comprises moving resources from one node to another node, and wherein resource dependencies determine an order that specific resources within a resource group are brought online or offline when the resource group is brought online or offline.

5. The method of claim 2, wherein the resource requirements are selected from the group consisting of:

CPU requirements, memory requirements, and disk performance requirements.

6. The method of claim 1, wherein the computer resources comprise computer nodes.

7. The method of claim 1, wherein the computer resources comprise data storage systems.

8. The method of claim 1, wherein the computer resources comprise storage area networks (SANs).

9. The method of claim 1, wherein the computer resources comprise computer networks and network devices.

10. The method of claim 1, wherein the cluster is part of a data center computer system.

11. The method of claim 1, wherein the failed resource is associated with a plurality of actions to be performed by the cluster in case of a failover, wherein upon an application failing in a cluster, another resource in the cluster executes the application from a point where a failure occurred.

12. A system comprising:

a plurality of interconnected computer resources;

logic using a hardware processor for representing the computer resources as nodes in a dependency graph, the nodes including a plurality of articulation points wherein removal of an articulation point due to a resource failure results in a disconnected dependency graph; and

logic for performing a failover for a resource group that a failed resource is a member if the failed resource corresponds to an articulation point, wherein all of the nodes in a cluster are aware of status of all resources on all other nodes, wherein if a failover of a remote resource only satisfies a portion of all resource requirements, then failover still occurs and the failover remote resource runs in a degraded mode,

wherein the resources comprise local resources and the failover is to a local resource if the failed resource does not affect all of the local resources, wherein upon a resource failing, analyzing the dependency graph to determine if an operational state of a primary node is reachable, and wherein resource dependencies determine an order that specific resources within a resource group are brought online or offline when the resource group is brought online or offline.

13. The system of claim 12, wherein the resources comprise remote resources, each resource is associated with a plurality of resource requirements, and the failover is to a remote resource if the remote resource can meet the resource requirements of the failed resource, wherein all of the nodes in a cluster are constantly aware of status of all resources on all other nodes, and failover comprises moving resources from one node to another node.

14. The system of claim 13, wherein a failover comprises failover of resources of an entire resource group that the failed resource is a member.

15. A computer program product comprising a computer readable hardware storage device having computer readable program code embodied therewith, the computer readable program code comprising:

computer readable program code configured to represent the computer resources as nodes in a dependency graph, the nodes including a plurality of articulation points wherein removal of an articulation point due to a resource failure results in a disconnected dependency graph; and

computer readable program code configured to perform a failover for a resource group that a failed resource is a member if the failed resource corresponds to an articulation point, wherein all of the nodes in a cluster are aware of status of all resources on all other nodes, wherein if a failover remote resource only satisfies a portion of all resource requirements, then failover still occurs and the failover remote resource runs in a degraded mode,

wherein the resources comprise local resources and the failover is to a local resource if the failed resource does not affect all of the local resources, wherein upon a resource failing, analyzing the dependency graph to determine if an operational state of a primary node is reachable, and wherein resource dependencies determine an order that specific resources within a resource group are brought online or offline when the resource group is brought online or offline.

16. The computer program product of claim 15, wherein the resources comprise remote resources, each resource is associated with a plurality of resource requirements, and the failover is to a remote resource if the remote resource can meet the resource requirements of the failed resource, wherein all of the nodes in a cluster are constantly aware of status of all resources on all other nodes, and wherein failover comprises moving resources from one node to another node.

17. The computer program product of claim 16, wherein comprises failover of resources of an entire resource group that the failed resource is a member.