Managing system power based on utilization statistics

• US 7,222,245 B2
• Filed: 04/26/2002
• Issued: 05/22/2007
• Est. Priority Date: 04/26/2002
• Status: Expired due to Fees

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Singh, Jitendra K.
• Primary Examiner(s): Perveen; Rehana
• Assistant Examiner(s): CHEN, TSE W
• Application Number:
  US10/132,931
• Publication Number:
  US 20030204758A1
• Time in Patent Office: 1,852 Days
• Field of Search: 718/105
• US Class Current: 713/300
• CPC Class Codes: G06F 1/3203 Power management, i.e. event-based initiation of a power-saving mode
• CPC Class Codes: G06F 1/324 by lowering clock frequency
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

The present invention, in various embodiments, provides techniques for managing system power. In one embodiment, system compute loads and/or system resources invoked by services running on the system consume power. To better manage power consumption, the spare capacity of a system resource is periodically measured, and if this spare capacity is outside a predefined range, then the resource operation is adjusted, e.g., the CPU speed is increased or decreased, so that the spare capacity is within the range. Further, the spare capacity is kept as close to zero as practical, and this spare capacity is determined based on the statistical distribution of a number of utilization values of the resources, which is also taken periodically. The spare capacity is also calculated based on considerations of the probability that the system resources are saturated. In one embodiment, to maintain the services required by a Service Level Agreement (SLA), a correlation between an SLA parameter and a resource utilization is determined. In addition to other factors and the correlation of the parameters, the spare capacity of the resource utilization is adjusted based on the spare capacity of the SLA parameter. Various embodiments include optimizing system performance before calculating system spare capacity, saving power for system groups or clusters, saving power for special conditions such as brown-out, high temperature, etc.

Claims

1. A method of managing power of a system, comprising the steps of:

identifying a resource consuming power of the system;

changing utilization of the resource affects a parameter;

representing a mean value of a plurality of utilization values of the resource by a first variable;

representing a mean value of a plurality of values of the parameter by a second variable;

determining a relationship between the first variable and the second variable;

for a selected value of the second variable, based on the relationship, selecting a corresponding value of the first variable, and

causing the resource to be utilized at a level corresponding to the selected value of the first variable;

wherein

a value of the first variable is calculated from a plurality of utilization values of the resource that are periodically obtained;

a value of the second variable is calculated from a plurality of values of the parameter that are periodically obtained;

associating the selected value of the second variable with a probability for this value to be below a predetermined value;

based on this probability and the selected value of the second variable, determining a new value of the second variable; and

selecting a new value of the first variable corresponding to the new value of the second variable.

2. The method of claim 1 wherein the relationship is based on a correlation coefficient between the first variable and the second variable;

a value of the correlation coefficient indicating a degree of correlation between the first variable and the second variable.

3. The method of claim 2 wherein the correlation coefficient is based on a mean of a plurality of utilization values of the resource and a mean of a plurality of values of the parameter.

4. The method of claim 1 wherein the relationship is derived from a probability distribution function of the first variable and the second variable.

5. The method of claim 1 wherein the relationship forms a trend line having a slope based on a correlation coefficient between the first variable and the second variable, the standard deviation of the first variable, and the standard deviation of the second variable.

6. The method of claim 5 wherein the slope is determined as

7. The method of claim 1 wherein the system provides computer services to users via a computer network.

8. The method of claim 1 wherein the resource is selected from a group consisting of memory, a CPU, a storage device, a bus, a peripheral device, a card for connecting accessories, and a peripheral device of the system.

9. The method of claim 1 wherein the parameter is selected from a group consisting of a time the system takes to respond to a request for a service, from a time a user requests a service to a time the user receives the service, and amount of information provided by the system in a time unit.

10. The method of claim 1 further comprising the step of adjusting operation of the resource to achieve the new value of the first variable.

11. The method of claim 1 wherein the probability corresponds to a third variable related to the standard deviation of the second variable.

12. The method of claim 11 wherein the new value of the second variable is calculated from the equation T=D−

(m*σ

_y) in which D representing the selected value of the second variable, T representing the new value of the second variable, m representing the third variable, and σ

_y representing the standard deviation of the second variable.

13. The method of claim 1 further comprising the steps of determining spare capacity of the second variable based on the probability, the new value for the second variable, the standard deviation of the second variable, and a value of the second variable representing a real time value of the second variable.

14. The method of claim 13 wherein Sσ

_y represents the spare capacity, T represents the new value for the second variable, σ

_y represents the standard deviation of the second variable, A represents the value of the second variable representing a real time of the second variable, m corresponds to the probability, and

15. The method of claim 13 further using the spare capacity of the second variable to determine a new value of the first variable corresponding to the new value of the second variable.

16. The method of claim 1 further comprising the step of periodically optimizing performance of the system.

17. The method of claim 1 further comprising the step of turning off a program running on the system upon meeting a condition that affects the power consumed by the system.

18. A method of managing power of a system, comprising the steps of:

identifying a resource consuming power of the system;

changing utilization of the resource affects a parameter;

periodically determining spare capacity of the resource; and

if the spare capacity is outside a predefined range, then adjusting operation of the resource that affects power consumed by the resource to adjust the spare capacity to within the predefined range;

wherein

a first variable represents a mean value of a plurality of utilization values of the resource;

a second variable represents a mean value of a plurality of values of the parameter;

the spare capacity of the resource is determined based on spare capacity of the parameter and a slope of trend line representing a relationship between the first variable and the second variable, wherein the spare capacity of the parameter is based on a first value of the second variable, a standard deviation of the second variable, a third variable representing a confidence level for the second variable not to exceed a second value, and a third value of the second variable representing a real time value of the second variable.

19. The method of claim 18 wherein Δ

x represents the spare capacity of the resource, Sσ

_y represents the spare capacity of the parameter, p represents the slope of the trend line, and

20. The method of claim 18 wherein Sσ

_y represent the spare capacity of the parameter, T represents the first value of the second variable, σ

_y represents the standard deviation of the second variable, m represents the third variable representing the confidence level for the second variable not to exceed the second value, and A representing the third value of the second variable representing a real time value of the second variable, and

21. The method of claim 18 wherein the trend line provides a slope depending on a correlation coefficient between the first variable and the second variable, the standard deviation of the first variable, and the standard deviation of the second variable.

22. The method of claim 21 wherein p represents the slope, σ

_x,y represents the correlation coefficient, σ

_x represents the standard deviation of the first variable, and σ

_y represents the standard deviation of the second variable, and

23. The method of claim 18 further comprising the step of consolidating program applications based on the spare capacity of the resource.

24. The method of claim 18 wherein the system provides computer services to users via a computer network.

25. The method of claim 18 further comprising the step of turning off a program application running on the system upon meeting a condition that affects the power consumed by the system.

26. A system comprising:

a resource consuming power of the system;

changing utilization of the resource affects a parameter;

a first variable representing a mean of a plurality of utilization values of the resource;

a second variable representing a mean of a plurality of values of the parameter;

means for periodically determining spare capacity of the resource based on spare capacity of the parameter and a relationship between the first variable and the second variable; and

means for adjusting operation of the system that affects power consumed by the resource in order to adjust the spare capacity of the resource to within a predefined range if the spare capacity is outside the predefined range, wherein the spare capacity of the parameter is based on a first value of the second variable, a standard deviation of the second variable, a third variable representing a confidence level for the second variable not to exceed a second value, and a third value of the second variable representing a real time value of the second variable.

27. The system of claim 26 being connected to a network.

28. The system of claim 26 wherein the parameter is determined based on a requirement of a client requesting a service from the system.