System and method for implementing an intelligent backup technique for cluster resources
View Patent ↗Method and system for implementing a backup in a cluster comprising a plurality of interconnected nodes, at least one of the nodes comprising a cluster resource manager (CRM), and at least one of the nodes comprising a policy engine (PE), the PE maintaining at least one dependency associated with at least a first resource executing on at least one of the nodes. For example, the method comprises, receiving by the CRM a backup request for the first resource from an administrator; responsive to the request, updating by the CRM the cluster configuration; communicating by the CRM to the PE a cluster status and the updated configuration; providing by the PE to the CRM an instruction sequence for carrying out the backup, the instruction sequence based on the dependency associated with the first resource; and responsive to the instruction sequence, carrying out by the CRM the backup of the first resource.
1. A method for implementing a backup technique in a cluster comprising a plurality of interconnected nodes, at least one of the nodes comprising a cluster resource manager, and at least one of the nodes comprising a policy engine, the policy engine maintaining at least one dependency associated with at least a first resource executing on at least one of the nodes, the method comprising:
receiving by the cluster resource manager a backup request for the first resource from an administrator of the cluster;
responsive to receipt of the backup request for the first resource, updating by the cluster resource manager the cluster configuration;
communicating by the cluster resource manager to the policy engine a cluster status and the updated cluster configuration;
providing by the policy engine to the cluster resource manager an instruction sequence for carrying out the requested backup, the instruction sequence being based on the dependency associated with the first resource; and
responsive to receipt of the instruction sequence, carrying out by the cluster resource manager the requested backup of the first resource;
wherein the policy engine maintains a dependency associated with a second resource with respect to the first resource;
wherein the second resource depends on the first resource; and
wherein the carrying out by the cluster resource manager the requested backup of the first resource further includes:
identifying by the cluster resource manager a failure of one of the first and second resources; and
responsive to the identification of the failure of the one of the first and second resources, performing by the cluster resource manager recovery of the failure using a dead man dependency technique.
2. The method of claim 1 wherein the carrying out by the cluster resource manager the requested backup of the first resource further includes freezing the first resource before backing up the first resource, and thawing the first resource after backing up the first resource.
3. The method of claim 2 ,
wherein the carrying out by the cluster resource manager the requested backup of the first resource further includes freezing the second resource before freezing the first resource, and thawing the first resource before thawing the second resource; and
wherein the carrying out by the cluster resource manager the requested backup of the first resource further includes freezing the second resource before backing up the second resource, and thawing the second resource after backing up the second resource.
4. The method of claim 1 , further including:
monitoring by the cluster resource manager the first resource; and
suspending by the cluster resource manager the monitoring the first resource during the carrying out by the cluster resource manager the requested backup of the first resource.
5. The method of claim 1 , further including:
monitoring by the cluster resource manager the first resource; and
altering by the cluster resource manager expectations of the monitoring during the carrying out by the cluster resource manager the requested backup of the first resource.
6. A system for implementing a backup technique in a cluster comprising a plurality of interconnected nodes, at least one of the nodes comprising a cluster resource manager, and at least one of the nodes comprising a policy engine, the policy engine maintaining at least one dependency with at least a first resource executing on at least one of the nodes, the system comprising:
means for receiving by the cluster resource manager a backup request for the first resource from an administrator of the cluster;
responsive to receipt of the backup request for the first resource, means for updating by the cluster resource manager the cluster configuration;
means for communicating by the cluster resource manager to the policy engine a cluster status and the updated cluster configuration;
means for providing by the policy engine to the cluster resource manager an instruction sequence for carrying out the requested backup, the instruction sequence being based on the dependency associated with the first resource; and
responsive to receipt of the instruction sequence, means for carrying out by the cluster resource manager the requested backup of the first resource;
wherein the policy engine maintains a dependency associated with a second resource with respect to the first resource;
wherein the second resource depends on the first resource; and
wherein the means for carrying out by the cluster resource manager the requested backup of the first resource further includes:
means for identifying by the cluster resource manager a failure of one of the first and second resources; and
responsive to the identification of the failure of the one of the first and second resources, means for performing by the cluster resource manager recovery of the failure;
wherein the means for performing by the cluster resource manager recovery of the failure includes means for performing a dead man dependency technique.
7. The system of claim 6 wherein the means for carrying out by the cluster resource manager the requested backup of the first resource further includes means for freezing the first resource before backing up the first resource, and means for thawing the first resource after backing up the first resource.
8. The system of claim 7 ,
wherein the carrying out by the cluster resource manager the requested backup of the first resource further includes freezing the second resource before freezing the first resource, and thawing the first resource before thawing the second resource; and
wherein the means for carrying out by the cluster resource manager the requested backup of the first resource further includes means for freezing the second resource before backing up the second resource, and means for thawing the second resource after backing up the second resource.
9. The system of claim 6 , further including:
means for monitoring by the cluster resource manager the first resource; and
during the carrying out by the cluster resource manager the requested backup of the first resource, one of means for suspending by the cluster resource manager the monitoring the first resource and altering by the cluster resource manager expectations of the monitoring.
10. A computer program product comprising non-transitory computer-readable medium having stored thereon instructions executable by a computer for implementing a backup technique in a cluster comprising a plurality of interconnected nodes, at least one of the nodes comprising a cluster resource manager, and at least one of the nodes comprising a policy engine, the policy engine maintaining at least one dependency associated with at least a first resource executing on at least one of the nodes, the instructions executable by the computer for causing the computer to:
receive by the cluster resource manager a backup request for the first resource from an administrator of the cluster;
responsive to receipt of the backup request for the first resource, update by the cluster resource manager the cluster configuration;
communicate by the cluster resource manager to the policy engine a cluster status and the updated cluster configuration;
provide by the policy engine to the cluster resource manager an instruction sequence for carrying out the requested backup, the instruction sequence being based on the dependency associated with the first resource; and
responsive to receipt of the instruction sequence, carry out by the cluster resource manager the requested backup of the first resource;
wherein a second resource depends on the first resource and that dependency is maintained by the policy engine; and
wherein the instructions executable by the computer for causing the computer to carry out by the cluster resource manager the requested backup of the first resource further include causing the computer to:
identify by the cluster resource manager a failure of one of the first and second resources; and
responsive to the identification of the failure of the one of the first and second resources, perform by the cluster resource manager recovery of the failure using a dead man dependency technique.
11. The computer program product of claim 10 wherein the instructions executable by the computer for causing the computer to carry out by the cluster resource manager the requested backup of the first resource further include causing the computer to freeze the first resource before backing up the first resource, and to thaw the first resource after backing up the second resource.
12. The computer program product of claim 11 ,
wherein the instructions executable by the computer for causing the computer to carry out by the cluster resource manager the requested backup of the first resource further include causing the computer to:
freeze the second resource before backing up the second resource and freezing the first resource;
thaw the second resource after backing up the first resource; and
thaw the first resource before thawing the second resource.
13. The computer program product of claim 10 ,
further including instructions executable by the computer for causing the computer to monitor by the cluster resource manager the first resource; and
wherein the instructions executable by the computer for causing the computer to carry out by the cluster resource manager the requested backup of the first resource further include causing the computer to one of suspend by the cluster resource manager the monitor by the cluster resource manager the first resource and alter by the cluster resource manager expectations of the monitoring.