Characterization and assignment of workload requirements to resources based on predefined categories of resource utilization and resource availability

• US 8,869,161 B2
• Filed: 11/18/2010
• Issued: 10/21/2014
• Est. Priority Date: 11/18/2010
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: DePetro, Matthew James
• Primary Examiner(s): Puente, Emerson
• Assistant Examiner(s): WU, BENJAMIN C
• Application Number:
  US12/949,494
• Publication Number:
  US 20120131593A1
• Time in Patent Office: 1,433 Days
• Field of Search: None
• US Class Current: 718/104
• CPC Class Codes: G06F 9/5088 involving task migration

Abstract

A method for managing computing resources includes generating a first workload metadata for a first workload, generating a second workload metadata for a second workload, and comparing the first workload and the second workload metadata against resource metadata. The method includes, based upon the comparison of workload metadata against resource metadata, identifying a potential conflict in resource requirements between the first workload and the computing resources available to the processing entity, and assigning the second workload for execution by one of the processing entities. The metadata characterize computing resources required by the associated workload. The first workload metadata is initially prioritized over the second workload metadata. The workloads are to be executed by processing entities. The resource metadata is available to the processing entities. The potential conflict in resource requirements does not exist between the resource requirements of the second workload and the computing resources of the processing entity.

Claims

1. A method for managing computing resources in a computer system, comprising:

generating a first workload metadata for a first workload;

generating a second workload metadata for a second workload, wherein;

the first workload metadata and the second workload metadata each specify computing resources required by the associated workload, wherein the required computing resources are specified by dividing each computing resource into discrete segments or categories of usage and identifying each computing resource required by the associated workload based on a specific category of usage;

the first workload is initially prioritized over the second workload based on the generated workload metadatas and current availability of computing resources on the computer system; and

the first workload and the second workload are scheduled to be assigned to and executed by one or more processing entities in a plurality of processing entities in the computer system;

generating a resource metadata of the computer system, wherein the resource metadata specifies computing resources available on a first processing entity and a second processing entity in the plurality of processing entities, wherein the available computing resources are specified by identifying each computing resource available on the first and second processing entities based on the same categories of usage from which computing resources required by the first and second workload are generated;

comparing the first workload metadata and the second workload metadata against the resource metadata;

based upon the comparison of the first and second workload metadata against the resource metadata, identifying a potential conflict between the first workload metadata and the resource metadata, wherein the potential conflict is identified by determining that any of the computing resources required by the first workload would cause any of the computing resources available on the first processing entity and the second processing entity to exceed its capacity;

reprioritizing the second workload for assignment over the first workload based on the identified potential conflict;

determining whether the first or the second processing entity has sufficient resources to execute the second workload by comparing the second workload metadata with the resource metadata; and

based on a determination that either the first or the second processing entity has sufficient resources to execute the second workload, assigning the second workload for execution by the first or second processing entity.

2. The method of claim 1, further comprising using one or more thresholds to divide each of the required computing resources into discrete segments or categories of usage.

3. The method of claim 1, wherein the first workload metadata and second workload metadata further comprise quantitative characterizations of the computing resources required by each workload.

4. The method of claim 3, wherein the potential conflict is identified if execution of the first workload by the first processing entity would cause the first processing entity to exceed one hundred percent of any of its available computing resources.

5. The method of claim 1, wherein the one or more processing entities comprises a pool of computing nodes.

6. The method of claim 1, wherein the one or more processing entities comprises a pool of processor cores.

7. The method of claim 1, wherein the one or more processing entities comprises a pool of processor threads.

8. The method of claim 1, wherein generating the workload metadata for a respective workload comprises:

loading the workload into a system for execution;

executing the workload with a first sample of data; and

measuring the computing resources used by the workload.

9. The method of claim 8, wherein generating the workload metadata for the respective workload further comprises:

executing the workload with a second sample of data;

measuring the computing resources used by the workload; and

averaging the computing resources used by the workload.

10. The method of claim 1, wherein:

the first workload metadata includes a first requirement of a computing resource;

the second workload metadata includes a second requirement of the computing resource;

and the second requirement exceeds the first requirement.

11. An article of manufacture for managing computing resources in a computer system, comprising:

a non-transitory computer readable medium; and

computer-executable instructions carried on the non-transitory computer readable medium, the instructions readable by a processor, the instructions, when read and executed, for causing the processor to;

generating a first workload metadata for a first workload;

generating a second workload metadata for a second workload, wherein;

the first workload metadata and the second workload metadata each specify computing resources required by the associated workload, wherein the required computing resources are specified by dividing each computing resource into discrete segments or categories of usage and identifying each computing resource required by the associated workload based on a specific category of usage;

the first workload is initially prioritized over the second workload based on the generated workload metadatas and current availability of computing resources on the computer system; and

the first workload and the second workload are scheduled to be assigned to and executed by one or more processing entities in a plurality of processing entities in the computer system;

generating a resource metadata of the computer system, wherein the resource metadata specifies computing resources available on a first processing entity and a second processing entity in the plurality of processing entities, wherein the available computing resources are specified by identifying each computing resource available on the first and second processing entities based on the same categories of usage from which computing resources required by the first and second workload are generated;

comparing the first workload metadata and the second workload metadata against the resource metadata;

based upon the comparison of the first and second workload metadata against the resource metadata, identifying a potential conflict between the first workload metadata and the resource metadata, wherein the potential conflict is identified by determining that any of the computing resources required by the first workload would cause any of the computing resources available on the first processing entity and the second processing entity to exceed its capacity;

reprioritizing the second workload for assignment over the first workload based on the identified potential conflict;

determining whether the first or the second processing entity has sufficient resources to execute the second workload by comparing the second workload metadata with the resource metadata; and

based on a determination that either the first or the second processing entity has sufficient resources to execute the second workload, assigning the second workload for execution by the first or second processing entity.

12. The article of claim 11, further comprising using one or more thresholds to divide each of the required computing resources into discrete segments or categories of usage.

13. The article of claim 11, wherein the first workload metadata and second workload metadata further comprise quantitative characterizations of the computing resources required by each workload.

14. The article of claim 13, wherein the potential conflict is identified by causing the processor to determine whether execution of the first workload by the first processing entity would cause the first processing entity to exceed one hundred percent of any of its available computing resources.

15. The article of claim 11, wherein the one or more processing entities comprises a pool of computing nodes.

16. The article of claim 11, wherein the one or more processing entities comprises a pool of processor cores.

17. The article of claim 11, wherein the one or more processing entities comprises a pool of processor threads.

18. The article of claim 11, wherein generating the workload metadata for a respective workload comprises:

loading the workload into a system for execution;

executing the workload with a first sample of data; and

measuring the computing resources used by the workload.

19. The article of claim 18, wherein generating the workload metadata for the respective workload further comprises:

executing the workload with a second sample of data;

measuring the computing resources used by the workload; and

averaging the computing resources used by the workload.

20. A system for managing computing resources in a computer system, comprising:

a processor;

a computer readable medium coupled to the processor; and

computer-executable instructions carried on the computer readable medium, the instructions readable by the processor, the instructions, when read and executed, for causing the processor to;

generating a first workload metadata for a first workload;

generating a second workload metadata for a second workload, wherein;

the first workload metadata and the second workload metadata each specify computing resources required by the associated workload, wherein the required computing resources are specified by dividing each computing resource into discrete segments or categories of usage and identifying each computing resource required by the associated workload based on a specific category of usage;

the first workload is initially prioritized over the second workload based on the generated workload metadatas and current availability of computing resources on the computer system; and

the first workload and the second workload are scheduled to be assigned to and executed by one or more processing entities in a plurality of processing entities in the computer system;

generating a resource metadata of the computer system, wherein the resource metadata specifies computing resources available on a first processing entity and a second processing entity in the plurality of processing entities, wherein the available computing resources are specified by identifying each computing resource available on the first and second processing entities based on the same categories of usage from which computing resources required by the first and second workload are generated;

comparing the first workload metadata and the second workload metadata against the resource metadata;

based upon the comparison of the first and second workload metadata against the resource metadata, identifying a potential conflict between the first workload metadata and the resource metadata, wherein the potential conflict is identified by determining that any of the computing resources required by the first workload would cause any of the computing resources available on the first processing entity and the second processing entity to exceed its capacity;

reprioritizing the second workload for assignment over the first workload based on the identified potential conflict;

determining whether the first or the second processing entity has sufficient resources to execute the second workload by comparing the second workload metadata with the resource metadata; and

based on a determination that either the first or the second processing entity has sufficient resources to execute the second workload, assigning the second workload for execution by the first or second processing entity.

21. The system of claim 20, further comprising using one or more thresholds to divide each of the required computing resources into discrete segments or categories of usage.

22. The system of claim 20, wherein the first workload metadata and second workload metadata further comprise quantitative characterizations of the computing resources required by each workload.

23. The system of claim 22, wherein the potential conflict is identified by causing the processor to determine whether execution of the first workload by the first processing entity would cause the first processing entity to exceed one hundred percent of any of its available computing resources.

24. The system of claim 20, wherein the one or more processing entities comprises a pool of computing nodes.

25. The system of claim 20, wherein the one or more processing entities comprises a pool of processor cores.

26. The system of claim 20, wherein the one or more processing entities comprises a pool of processor threads.

27. The system of claim 20, wherein generating the workload metadata for a respective workload comprises:

loading the workload into a system for execution;

executing the workload with a first sample of data; and

measuring the computing resources used by the workload.

28. The system of claim 20, wherein generating the workload metadata for the respective workload further comprises:

executing the workload with a second sample of data;

measuring the computing resources used by the workload; and

averaging the computing resources used by the workload.