Multi-variate computer resource allocation

• US 9,432,443 B1
• Filed: 01/31/2007
• Issued: 08/30/2016
• Est. Priority Date: 01/31/2007
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Inventors: Coffey, James Mark, Herington, Daniel Edward
• Primary Examiner(s): Donabed, Ninos
• Application Number:
  US11/669,708
• Time in Patent Office: 3,499 Days
• Field of Search: 709/226
• US Class Current: 1/1
• CPC Class Codes: G06F 15/173 using an interconnection network, e.g. matrix, shuffle, pyramid, star, snowflake
• CPC Class Codes: G06F 2209/501 Performance criteria
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: H04L 41/0813 characterised by the conditions triggering a change of settings
• CPC Class Codes: H04L 41/50 Network service management, e.g. ensuring proper service fulfilment according to agreements
• CPC Class Codes: H04L 43/0817 by checking functioning
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46

Abstract

The present invention provides a method for multi-variate allocation of hardware resources to workloads on a computer system as a function of load and at least one of performance and utilization.

Claims

1. A computer-implemented multi-variate workload management method comprising:

tracking response time and transaction rates as a function of time for respective workloads over a duration in which hardware resource allocations to said workloads have changed to obtain allocation histories to determine impacts of transaction rates on the responsiveness of response time to changes in resource allocations;

determining hardware requirements for respective workloads based at least in part on said impacts of transaction rates and on management policies, said management policies specifying criteria for allocating said resources among said workloads, said criteria include required response times for respective workloads and arbitration criteria for arbitrating among competing hardware resource requirements for said workloads when not all of the hardware resource requirements can be met;

determining an allocation of hardware resources to said workloads based on said hardware requirements and said management policies; and

implementing said allocation on said computer system.

2. A method as recited in claim 1 wherein:

said transaction rates are numbers of transactions being initiated per unit time; and

said response times are average response times associated with said transactions.

3. A computer system comprising:

hardware resources;

plural nodes among which said hardware resources can be allocated;

plural workloads running on said plural nodes;

management policies for specifying criteria for allocating said resources among said workloads, said criteria including required response times for respective workloads and arbitration criteria for arbitrating among competing hardware resource requirements for said workloads when not all of the hardware resource requirements can be met;

allocation histories representing changes in response time as functions of time, resource allocations, and transaction rates; and

a workload manager to;

determine hardware requirements for respective workloads based at least in part on said management policies and an impact of said transaction rates on responsiveness of response times to changes in resource allocations, said impact being determined from said allocation histories; and

allocate hardware resources according to the determined hardware requirements.

4. A computer system as recited in claim 3 wherein:

said transaction rates are numbers of transactions being initiated per unit time; and

said response times are average response times associated with said transactions.