Predictive control for resource entitlement

• US 8,732,307 B1
• Filed: 07/25/2006
• Issued: 05/20/2014
• Est. Priority Date: 07/25/2006
• Status: Active Grant

• Current Assignee: Valtrus Innovations Limited (f/k/a Dolya Holdco 9 Limited) (Key Patent Innovations Limited)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Zhu, Xiaoyun, Xu, Wei, Singhal, Sharad, Wang, Zhikui
• Primary Examiner(s): THIAW, CATHERINE B
• Application Number:
  US11/492,354
• Time in Patent Office: 2,856 Days
• Field of Search: 709221-226, 709235-237, 370235-237, 718/104
• US Class Current: 709/226
• CPC Class Codes: G06F 2209/5019 Workload prediction
• CPC Class Codes: G06F 2209/504 Resource capping
• CPC Class Codes: G06F 9/5061 Partitioning or combining of resources
• CPC Class Codes: G06F 9/5072 Grid computing
• CPC Class Codes: G06F 9/5077 Logical partitioning of resources; Management or configuration of virtualized resources specific details on emulation or internal functioning of virtual machines G06F9/455
• CPC Class Codes: G06F 9/52 Program synchronisation; Mutual exclusion, e.g. by means of semaphores
• CPC Class Codes: H04L 47/822 Collecting or measuring resource availability data
• CPC Class Codes: H04L 47/823 Prediction of resource usage
• CPC Class Codes: H04L 47/83 based on usage prediction
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

A resource consumption of a resource compartment for a time interval is predicted. A resource entitlement for the resource compartment for the time interval is determined based on the predicted resource consumption.

Claims

1. A method for controlling resource entitlement in a resource-on-demand system, the method comprising:

predicting, by a resource allocator in a computer system, a resource consumption of a resource compartment for a time interval;

determining, by the resource allocator, a resource entitlement for the resource compartment for the time interval such that the resource entitlement is equal to the predicted resource consumption divided by a target resource utilization; and

sending the resource entitlement to a resource scheduler to allocate the resource entitlement to the resource compartment.

2. The method of claim 1, wherein predicting a resource consumption of a resource compartment for a time interval further comprises:

determining resource consumptions for previous time intervals for the resource compartment; and

predicting the resource consumption for the resource compartment for the time interval based on the resource consumptions for the previous time intervals.

3. The method of claim 2, wherein predicting the resource consumption for the resource compartment for the time interval based on the resource consumptions for the previous time intervals further comprises:

predicting the resource consumption for the resource compartment for the future time interval using one of an auto-regressive model, ANOVA decomposition, and a multi-pulse model.

4. The method of claim 1, further comprising:

measuring a resource consumption for the resource compartment for the time interval; and

storing the resource consumption for the time interval.

5. The method of claim 1, further comprising:

using the stored resource consumption and other stored resource consumptions for the resource compartment to predict a resource consumption and determine a resource entitlement for a future time interval.

6. A predictive control system for controlling resource entitlement for a resource compartment comprising:

a resource scheduler operable to schedule a resource entitlement for the resource compartment;

a sensor operable to determine resource consumption of the resource compartment; and

a predictive controller operable to predict a resource consumption of a resource compartment for a time interval and determine a resource entitlement for the resource compartment for the time interval such that the resource entitlement is equal to the predicted resource consumption divided by a target resource utilization, and the predictive controller is further operable to send the resource entitlement to the resource scheduler to allocate the resource entitlement to the resource compartment.

7. The system of claim 6, wherein the resource entitlement is an allocation of a resource comprising one of CPU cycles, network bandwidth, and storage space.

8. The system of claim 6, wherein the predictive controller comprises a predictor operable to predict the resource consumption of the resource compartment for the time interval based on measured consumptions for previous time intervals.

9. The system of claim 6, wherein the resource compartment comprises one of a process group and a virtual machine.

10. The system of claim 6, wherein the sensor measures the resource consumption for the resource compartment for the time interval after the resource entitlement is allocated to the resource compartment, and the measured resource consumption is stored for predicting resource consumption for future time intervals.

11. A computer readable storage device upon which is embedded programmed instructions which when executed by a processor will cause the processor to perform a method for controlling resource entitlement in a resource-on-demand system, the method comprising:

predicting a resource consumption of a resource compartment for a time interval;

determining a resource entitlement for the resource compartment for the time interval such that the resource entitlement is equal to the predicted resource consumption divided by a target resource utilization; and

sending the resource entitlement to a resource scheduler to allocate the resource entitlement to the resource compartment.

12. The computer readable storage device of claim 11, wherein predicting a resource consumption of a resource compartment for a time interval further comprises:

determining resource consumptions for previous time intervals for the resource compartment; and

predicting the resource consumption for the resource compartment for the time interval based on the resource consumptions for the previous time intervals.

13. The computer readable storage device of claim 12, wherein predicting the resource consumption for the resource compartment for the time interval based on the resource consumptions for the previous time intervals further comprises:

predicting the resource consumption for the resource compartment for the future time interval using one of an auto-regressive model, ANOVA decomposition, and a multi-pulse model.

14. The computer readable storage device of claim 11, wherein the method further comprises:

measuring a resource consumption for the resource compartment for the time interval; and

storing the resource consumption for the time interval.

15. The computer readable storage device of claim 14, wherein method further comprises:

using the stored resource consumption and other stored resource consumptions for the resource compartment to predict a resource consumption and determine a resource entitlement for a future time interval.