System and method for determining allocation of resource access demands to different classes of service based at least in part on permitted degraded performance

• US 8,046,765 B2
• Filed: 07/25/2006
• Issued: 10/25/2011
• Est. Priority Date: 07/25/2006
• 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: Rolia, Jerome, Cherkasova, Ludmila
• Primary Examiner(s): An; Meng
• Assistant Examiner(s): KUMABE, BLAKE K
• Application Number:
  US11/492,376
• Publication Number:
  US 20080028409A1
• Time in Patent Office: 1,918 Days
• Field of Search: 718/104
• US Class Current: 718/104
• CPC Class Codes: G06F 2209/5011 Pool
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources

Abstract

A method comprises receiving into a planning tool a representative workload for a consumer. The method further comprises receiving into the planning tool quality of service desires of the consumer which define permitted degraded performance. In certain embodiments, the permitted degraded performance is time-limited wherein demands of the representative workload may exceed a pre-defined utilization constraint for at least one resource servicing the demands for no more than a pre-defined amount of contiguous time. The planning tool determines an allocation of demand of the consumer for each of a plurality of different classes of service (COSs). In certain embodiments, a first COS provides guaranteed resource access for servicing demand allocated thereto, and a second COS provides non-guaranteed resource access for servicing demand allocated thereto. In certain embodiments, the allocation of demand to the different COSs may be determined for both a normal mode and a failure mode of operation.

Claims

1. A method comprising:

receiving into a planning tool executed in a computer system a representative workload for a consumer;

receiving into the planning tool quality of service objectives of the consumer which define permitted degraded performance, wherein the quality of service objectives which define permitted degraded performance include time information specifying a limit on a length of time over which the degraded performance is permissible, wherein said time information specifies a limit on a length of contiguous time that demands of the representative workload may exceed a pre-defined utilization constraint for at least one resource servicing the demands of the representation workload; and

determining, by the planning tool based on the quality of service objectives which define permitted degraded performance, an allocation of demands of the consumer across a plurality of different classes of service (COSs).

2. The method of claim 1 wherein the permitted degraded performance comprises:

performance of the at least one resource in servicing the demands of the representative workload, wherein utilization of the at least one resource for servicing the demands of the representative workload exceeds the pre-defined utilization constraint.

3. The method of claim 2 further comprising:

receiving into the planning tool information defining the utilization constraint.

4. The method of claim 1 wherein said receiving into the planning tool the quality of service objectives of the consumer which define permitted degraded performance comprises:

receiving into the planning tool first quality of service objectives of the consumer which define permitted degraded performance for a normal mode of operation in which all resources in a pool of resources to which the representative workload is assigned are operational; and

receiving into the planning tool second quality of service objectives of the consumer which define permitted degraded performance for a failure mode of operation in which at least one of the resources in the pool of resources to which the representative workload is assigned is not operational.

5. The method of claim 4 wherein said determining, by the planning tool, the allocation of the demands of the consumer across the plurality of different COSs comprises:

determining, by the planning tool, a first allocation of the demands of the consumer across the plurality of different COSs for the normal mode of operation; and

determining, by the planning tool, a second allocation of the demands of the consumer across the plurality of different COSs for the failure mode of operation.

6. The method of claim 1 further comprising:

outputting, by said planning tool, said determined allocation, wherein said outputting comprises at least one selected from the group consisting of presenting to a display, outputting to data storage, and communicating to an application.

7. The method of claim 1 wherein said plurality of different COSs comprise:

a first COS having guaranteed resource access for servicing demand allocated to said first COS; and

a second COS having non-guaranteed resource access for servicing demand allocated to said second COS.

8. The method of claim 7 further comprising:

receiving, by said planning tool, input specifying a particular parameter, wherein said second COS has a resource access probability of no worse than said particular parameter.

9. The method of claim 1 further comprising:

receiving, by said planning tool, resource access quality of service (QoS) commitments which define a permitted degree of overbooking for resources in a resource pool to which the representative workload is assigned.

10. The method of claim 9 wherein said receiving, by said planning tool, the resource access QoS commitments comprises:

receiving, by said planning tool, information defining a resource access probability for a non-guaranteed COS; and

receiving, by said planning tool, information specifying a deadline by which demands of the representative workload that are not satisfied upon request are to be satisfied.

11. The method of claim 9 wherein the determining, by the planning tool, the allocation of demands of the consumer across the plurality of different COSs comprises:

determining the allocation of the demands of the consumer across the plurality of different COSs that satisfies both the received quality of service objectives of the consumer and the received resource access QoS commitments.

12. The method of claim 1, wherein the representative workload is an estimation of a future workload for the consumer, and wherein the representative workload is based on a historical workload for the consumer.

13. A method comprising:

receiving into a planning tool executed in a computer system a representative workload for a consumer;

receiving into the planning tool quality of service objectives of the consumer which define permitted degraded performance, wherein the quality of service objectives which define permitted degraded performance include time information specifying a limit on a length of time over which the degraded performance is permissible, and wherein the quality of service objectives further include information specifying a permitted percentage of demands in the representative workload that may exceed a pre-defined utilization constraint for at least one resource servicing the demands of the representative workload; and

determining, by the planning tool based on the quality of service objectives which define permitted degraded performance, an allocation of demands of the consumer across a plurality of different classes of service (COSs).

14. A non-transitory computer-readable medium storing code executable in a computer system, the code comprising:

code for receiving a representative workload for a consumer;

code for receiving information defining desired performance for servicing the representative workload of the consumer, wherein the information comprises

a) information defining a constraint on utilization of at least one resource in servicing demands of the representative workload, and

b) information defining a permissible time-limited degraded performance specifying a limit on a length of contiguous time during which the defined constraint on utilization of the at least one resource can be exceeded in servicing demands of the representative workload; and

code for determining a breakpoint for partitioning resource demands of said representative workload across a plurality of different classes of service (COSs) in compliance with said information defining the desired performance.

15. The computer-readable medium of claim 14, wherein said plurality of different COSs comprise:

a first COS having guaranteed resource access for servicing demand allocated to said first COS; and

a second COS having non-guaranteed resource access for servicing demand allocated to said second COS.

16. The computer-readable medium of claim 14 wherein said code for receiving information defining desired performance for servicing the representative workload of the consumer comprises:

a) code for receiving said information defining desired performance for a normal mode of operation in which all resources in a pool of resources to which the representative workload is assigned are operational; and

b) code for receiving said information defining desired performance for a failure mode of operation in which at least one of the resources in the pool of resources to which the representative workload is assigned is not operational; and

wherein said code for determining the breakpoint comprises;

a) code for determining a first breakpoint for partitioning resource demands of said representative workload between the plurality of different COSs in compliance with said information defining the desired performance for the normal mode of operation; and

b) code for determining a second breakpoint for partitioning resource demands of said representative workload between the plurality of different COSs in compliance with said information defining the desired performance for the failure mode of operation.

17. The computer-readable medium of claim 14, wherein the specified length of time is a contiguous length of time.

18. A system comprising:

at least one central processing unit (CPU);

a scheduler for at least one resource, said scheduler executable on the at least one CPU to schedule access to the at least one resource for servicing demands of a consumer based at least in part on a plurality of different classes of service (COSs) for the demands; and

a planning tool executable on the at least one CPU to;

receive a representative workload of the consumer;

receive information defining desired performance for servicing the representative workload of the consumer, wherein the information comprises

a) information defining a constraint on utilization of the at least one resource in servicing demands of the representative workload, and

b) information defining a permissible time-limited degraded performance specifying a limit on a length of contiguous time during which the defined constraint on utilization of the at least one resource can be exceeded in servicing demands of the representative workload; and

determine at least one breakpoint for partitioning resource demands of the representative workload across the plurality of different COSs in compliance with said information defining the desired performance.

19. The system of claim 18 wherein the determined at least one breakpoint is used by the scheduler for servicing the demands of the consumer.

20. The system of claim 18 wherein said plurality of different COSs comprise:

a first COS having guaranteed resource access for servicing demand allocated to said first COS; and

a second COS having non-guaranteed resource access for servicing demand allocated to said second COS.

21. The system of claim 18 wherein the information defining the permissible time-limited degraded performance specifying the length of time during which the defined constraint on utilization of the at least one resource can be exceeded in servicing demands of the representative workload comprises:

information defining a permitted percentage of demands in the representative workload that may exceed the defined constraint on utilization of the at least one resource; and

time information defining the length of time.

22. The system of claim 18, wherein the specified length of time is a contiguous length of time.