Terminating a non-clustered workload in response to a failure of a system with a clustered workload

• US 7,653,833 B1
• Filed: 10/31/2006
• Issued: 01/26/2010
• Est. Priority Date: 10/31/2006
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Miller, Troy Don, Crawford, Isom Lawrence Jr.
• Primary Examiner(s): McCarthy; Christopher S
• Application Number:
  US11/590,321
• Time in Patent Office: 1,183 Days
• Field of Search: None
• US Class Current: 714/13
• CPC Class Codes: G06F 11/1438 Restarting or rejuvenating
• CPC Class Codes: G06F 11/203 using migration
• CPC Class Codes: G06F 11/2035 without idle spare hardware
• CPC Class Codes: G06F 2201/815 Virtual middleware or OS functionality using virtual machines to implement generic software techniques for error detection or fault masking G06F11/1484

Abstract

The present invention provides for check-pointing an non-clustered workload to make room for a clustered workload that was running on a computer system that has suffered a hardware failure.

Claims

1. An automated method comprising:

clustering first and second workloads, said first and second workloads being on respective first and second computer systems;

characterizing a third workload as non-clustered, said third workload being on said second computer system;

detecting a system failure on said first computer system;

determining whether said second computer system has sufficient resources to allow the clustered workload running on said first computer system to migrate to the second computer system;

if said resources are not sufficient,

check-pointing said non-clustered workload on said second computer system at a checkpoint state, and

migrating said workload from said first computer system to said second computer system.

2. A method as recited in claim 1 wherein, if said resources are sufficient, migrating said first clustered workload to said second computer system without check-pointing said non-clustered workload.

3. A method as recited in claim 1 further comprising, after checkpointing said non-clustered workload, migrating it to a third computer system and resuming its operation from said checkpoint state.

4. A method as recited in claim 3 further comprising, after said detecting said failure, migrating a nonclustered workload that was running on first computer system before said failure to said third computer system.

5. A method as recited in claim 1 further comprising, after repairing said first computer system, returning said first computer workload to said first computer system.

6. A clustered computer network comprising:

a first computer system;

a second computer system;

first, second, and third workloads, said first workload running on said first computer system, said second and third workloads running on said second computer system; and

a workload manager, said workload manager assigning said first and second workloads to a common cluster and said third workload to no cluster so that, in the event said first system fails, said third workload is check-pointed and said first workload is migrated to said second computer system.

7. A clustered computer network as recited in claim 6 wherein said workload manager migrates said third workload to a third computer system after said third workload is check-pointed so that it resumes operation from its check-pointed state.

8. A clustered computer network as recited in claim 7 wherein, after said first computer system is repaired, said workload manager restores said workload so that it runs on said first computer system.