Computer resource allocation as a function of demand type

• US 8,156,502 B1
• Filed: 06/08/2007
• Issued: 04/10/2012
• Est. Priority Date: 06/08/2007
• 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: Blanding, William H.
• Primary Examiner(s): An, Meng
• Assistant Examiner(s): LEE, TAMMY EUNHYE
• Application Number:
  US11/760,635
• Time in Patent Office: 1,768 Days
• Field of Search: None
• US Class Current: 718/104
• CPC Class Codes: G06F 2209/508 Monitor
• CPC Class Codes: G06F 9/50 Allocation of resources, e.g. of the central processing unit [CPU]

Abstract

A method of allocating computer resources to workloads involves monitoring utilization of computer resources by workloads to generate utilization as a function of time data. The utilization data is used to classify workloads into demand types as a function of the utilization data. A workload manager plans and implements an allocation of computer resources to the workloads in part as a function of the demand types of the workloads.

Claims

1. A method of allocating computer resources to workloads comprising:

monitoring utilization of computer resources by workloads and/or components of workloads to generate utilization data of one or more computer resource types as a function of time;

classifying said workloads into demand types as a function of said utilization data;

planning an allocation of said computer resources to said workloads in part as a function of the demand types of said workloads;

implementing said allocation; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said implementing involves transferring resources into or out of said partitions.

2. A method as recited in claim 1 wherein said implementing involves creating partitions to which said workloads may be assigned and assigning said workloads to said partitions according to demand type.

3. A method as recited in claim 1 wherein said implementing involves migrating workloads between partitions.

4. A method as recited in claim 1 wherein said planning involves grouping workloads by demand type to form workload groups to which some resources are collectively allocated.

5. A method as recited in claim 1 wherein at least one of said workloads includes at least one operating system.

6. A method as recited in claim 1 wherein at least two of said workloads are running on the same operating system instance.

7. A method as recited in claim 1 wherein said demand types are selected from a set including at least a full demand type with relatively low utilization variance for a given allocation of resources and an intermittent demand type having a relatively high utilization variance with a peak utilization in an allocation period more than twice its average utilization for the same allocation period.

8. A method of allocating computer resources to workloads comprising:

monitoring utilization of computer resources by workloads and/or components of workloads to generate utilization data of one or more computer resource types as a function of time;

classifying said workloads into demand types as a function of said utilization data;

planning an allocation of said computer resources to said workloads in part as a function of the demand types of said workloads;

implementing said allocation; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said implementing involves transferring workloads into or out of said partitions.

9. A method of allocating computer resources to workloads comprising:

monitoring utilization of computer resources by workloads and/or components of workloads to generate utilization data of one or more computer resource types as a function of time;

classifying said workloads into demand types as a function of said utilization data;

planning an allocation of said computer resources to said workloads in part as a function of the demand types of said workloads;

implementing said allocation; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions, and said implementing involves transferring workload components into or out of said partitions and transferring resources into or out of said partitions.

10. A method of allocating computer resources to workloads comprising:

monitoring utilization of computer resources by workloads and/or components of workloads to generate utilization data of one or more computer resource types as a function of time;

classifying said workloads into demand types as a function of said utilization data;

planning an allocation of said computer resources to said workloads in part as a function of the demand types of said workloads;

implementing said allocation; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to existing partitions and/or to newly created partitions, and said implementing involves creating said partitions as necessary and transferring workload components into or out of said partitions and transferring resources into or out of said partitions.

11. A computer system comprising:

computer resources;

workloads running on said resources;

and a computer-executable workload manager for determining workload demand types and allocating said resources to said workloads in part as a function of demand type, said workload manager including a utilization monitor for monitoring resource utilization by said workloads to obtain utilization-over-time data for each of said workloads, said workload manager classifying said workloads by demand type as a function of said utilization-over-time data; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said allocating involves transferring resources into or out of said partitions.

12. A system as recited in claim 11 wherein said workload manager defined demand types include a full demand type which is characterized by relatively low utilization variance for a given resource allocation and an intermittent demand type which is characterized by a relatively high utilization variance.

13. A system as recited in claim 12 wherein said workloads are processes, said workload manager separating said processes by type by creating a separate partition for each demand type.

14. A system as recited in claim 13 wherein said workload manager assigns plural processes of said intermittent demand to the same process partition.

15. A system as recited in claim 11 wherein each of said workloads includes an operating system instance.

16. A system as recited in claim 15 wherein said workload manager allocates resources collectively to groups of said workloads of the same demand type.

17. A system as recited in claim 11 wherein said workload manager creates virtual machines and allocates resources to demand-type groups of said virtual machines having the same demand type.

18. A computer system comprising:

computer resources;

workloads running on said resources;

and a computer-executable workload manager for determining workload demand types and allocating said resources to said workloads in part as a function of demand type, said workload manager including a utilization monitor for monitoring resource utilization by said workloads to obtain utilization-over-time data for each of said workloads, said workload manager classifying said workloads by demand type as a function of said utilization-over-time data; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said allocating involves transferring workloads into or out of said partitions.

19. A computer system comprising:

computer resources;

workloads running on said resources;

and a computer-executable workload manager for determining workload demand types and allocating said resources to said workloads in part as a function of demand type, said workload manager including a utilization monitor for monitoring resource utilization by said workloads to obtain utilization-over-time data for each of said workloads, said workload manager classifying said workloads by demand type as a function of said utilization-over-time data; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions, and said allocating involves transferring workload components into or out of said partitions and transferring resources into or out of said partitions.

20. A computer system comprising:

computer resources;

workloads running on said resources;

and a computer-executable workload manager for determining workload demand types and allocating said resources to said workloads in part as a function of demand type, said workload manager including a utilization monitor for monitoring resource utilization by said workloads to obtain utilization-over-time data for each of said workloads, said workload manager classifying said workloads by demand type as a function of said utilization-over-time data; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to existing partitions and/or to newly created partitions, and said allocating involves creating said partitions as necessary and transferring workload components into or out of said partitions and transferring resources into or out of said partitions.

21. Non-transitory computer-readable media comprising a computer-executable workload manager including a utilization monitor for monitoring utilization and generating utilization-over-time data, said workload manager classifying workloads according to demand type as a function of said data, said workload manager allocating computer resources to said workloads in part as a function of demand type;

and wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said allocating involves transferring resources into or out of said partitions.

22. Media as recited in claim 21 wherein the demand type for a workload is selected from a set of demand types including a full utilization type and an intermittent type.

23. Media as recited in claim 22 wherein said workload manager creates partitions so that workloads of the same demand type are in the same partition.

24. Media as recited in claim 23 where the variance of said intermittent type is greater than that of said full utilization type.

25. Media as recited in claim 21 wherein at least some of said workloads include operating system instances.

26. Media as recited in claim 21 wherein at least some of said workloads are intra-operating system processes.

27. Non-transitory computer-readable media comprising a computer-executable workload manager including a utilization monitor for monitoring utilization and generating utilization-over-time data, said workload manager classifying workloads according to demand type as a function of said data, said workload manager allocating computer resources to said workloads in part as a function of demand type;

and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said allocating involves transferring workloads into or out of said partitions.

28. Non-transitory computer-readable media comprising a computer-executable workload manager including a utilization monitor for monitoring utilization and generating utilization-over-time data, said workload manager classifying workloads according to demand type as a function of said data, said workload manager allocating computer resources to said workloads in part as a function of demand type;

and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions, and said allocating involves transferring workload components into or out of said partitions and transferring resources into or out of said partitions.

29. Non-transitory computer-readable media comprising a computer-executable workload manager including a utilization monitor for monitoring utilization and generating utilization-over-time data, said workload manager classifying workloads according to demand type as a function of said data, said workload manager allocating computer resources to said workloads in part as a function of demand type;

and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to existing partitions and/or to newly created partitions, and said allocating involves creating said partitions as necessary and transferring workload components into or out of said partitions and transferring resources into or out of said partitions.

30. A computer-implemented method comprising:

monitoring utilization of hardware computing resources by workloads to obtain respective utilization histories for said workloads;

classifying said workloads according to demand type as a function of their respective utilization histories so that plural workloads are characterized by an intermittent demand type in that they have relatively high utilization variances and plural other workloads are characterized by a constant demand type in that they have relatively low utilization variance;

grouping said workloads by demand types to create demand-type groups of workloads;

allocating said hardware computing resources to said groups at least in part according to their respective demand types; and

said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said allocating involves transferring resources into or out of said partitions.

31. A method as recited in claim 30 wherein:

said monitoring involves continuing to monitor after said workloads have been classified a first time to obtain modified utilization histories for said workloads; and

said classifying involves reclassifying at least one of said workloads according to its modified utilization history so that the demand type with which it is characterized changes from one demand type to another demand type.

32. A computer-implemented method comprising:

monitoring utilization of hardware computing resources by workloads to obtain respective utilization histories for said workloads;

classifying said workloads according to demand type as a function of their respective utilization histories so that plural workloads are characterized by an intermittent demand type in that they have relatively high utilization variances and plural other workloads are characterized by a constant demand type in that they have relatively low utilization variance;

grouping said workloads by demand types to create demand-type groups of workloads;

allocating said hardware computing resources to said groups at least in part according to their respective demand types; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said allocating involves transferring workloads into or out of said partitions.

33. A computer-implemented method comprising:

monitoring utilization of hardware computing resources by workloads to obtain respective utilization histories for said workloads;

classifying said workloads according to demand type as a function of their respective utilization histories so that plural workloads are characterized by an intermittent demand type in that they have relatively high utilization variances and plural other workloads are characterized by a constant demand type in that they have relatively low utilization variance;

grouping said workloads by demand types to create demand-type groups of workloads;

allocating said hardware computing resources to said groups at least in part according to their respective demand types; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions, and said allocating involves transferring workload components into or out of said partitions and transferring resources into or out of said partitions.

34. A computer-implemented method comprising:

monitoring utilization of hardware computing resources by workloads to obtain respective utilization histories for said workloads;

classifying said workloads according to demand type as a function of their respective utilization histories so that plural workloads are characterized by an intermittent demand type in that they have relatively high utilization variances and plural other workloads are characterized by a constant demand type in that they have relatively low utilization variance;

grouping said workloads by demand types to create demand-type groups of workloads;

allocating said hardware computing resources to said groups at least in part according to their respective demand types; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to existing partitions and/or to newly created partitions, and said allocating involves creating said partitions as necessary and transferring workload components into or out of said partitions and transferring resources into or out of said partitions.

35. A system comprising non-transitory computer-readable storage media encoded with code that, when executed by a processor, causes:

monitoring utilization of hardware computing resources by workloads to obtain respective utilization histories for said workloads;

classifying said workloads according to demand type as a function of their respective utilization histories so that plural workloads are characterized by an intermittent demand type in that they have relatively high utilization variances and plural other workloads are characterized by a constant demand type in that they have relatively low utilization variance;

grouping said workloads by demand types to create demand-type groups of workloads;

allocating said hardware computing resources to said groups at least in part according to their respective demand types; and

said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said allocating involves transferring resources into or out of said partitions.

36. A system as recited in claim 35 wherein:

said monitoring involves continuing to monitor have said workloads have been classified a first time to obtain modified utilization histories for said workloads; and

said classifying involves reclassifying at least one of said workloads according to its modified utilization history so that the demand type with which it is characterized changes from one demand type to another demand type.

37. A system as recited in claim 36 further comprising said processor.

38. A system comprising non-transitory computer-readable storage media encoded with code that, when executed by a processor, causes:

monitoring utilization of hardware computing resources by workloads to obtain respective utilization histories for said workloads;

classifying said workloads according to demand type as a function of their respective utilization histories so that plural workloads are characterized by an intermittent demand type in that they have relatively high utilization variances and plural other workloads are characterized by a constant demand type in that they have relatively low utilization variance;

grouping said workloads by demand types to create demand-type groups of workloads;

allocating said hardware computing resources to said groups at least in part according to their respective demand types; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions and said allocating involves transferring workloads into or out of said partitions.

39. A system comprising non-transitory computer-readable storage media encoded with code that, when executed by a processor, causes:

monitoring utilization of hardware computing resources by workloads to obtain respective utilization histories for said workloads;

classifying said workloads according to demand type as a function of their respective utilization histories so that plural workloads are characterized by an intermittent demand type in that they have relatively high utilization variances and plural other workloads are characterized by a constant demand type in that they have relatively low utilization variance;

grouping said workloads by demand types to create demand-type groups of workloads;

allocating said hardware computing resources to said groups at least in part according to their respective demand types; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to partitions, and said allocating involves transferring workload components into or out of said partitions and transferring resources into or out of said partitions.

40. A system comprising non-transitory computer-readable storage media encoded with code that, when executed by a processor, causes:

monitoring utilization of hardware computing resources by workloads to obtain respective utilization histories for said workloads;

classifying said workloads according to demand type as a function of their respective utilization histories so that plural workloads are characterized by an intermittent demand type in that they have relatively high utilization variances and plural other workloads are characterized by a constant demand type in that they have relatively low utilization variance;

grouping said workloads by demand types to create demand-type groups of workloads;

allocating said hardware computing resources to said groups at least in part according to their respective demand types; and

wherein said workloads are grouped into demand-type groups, said demand-type groups are assigned to existing partitions and/or to newly created partitions, and said allocating involves creating said partitions as necessary and transferring workload components into or out of said partitions and transferring resources into or out of said partitions.