Method and apparatuses for determining a leak of resource and predicting usage of resource

Patent metadata

Field	Value
Publication number	US-9846601-B2
Application number	US-201514922595-A
Country	US
Kind code	B2
Filing date	Oct 26, 2015
Priority date	May 21, 2013
Publication date	Dec 19, 2017
Grant date	Dec 19, 2017

Abstract

A method and an apparatus for determining a leak of a program running resource are disclosed that relate to the field of computer applications. The method for predicting a usage condition of a program running resource includes collecting program running resource usage at least once within each program running resource usage period; decomposing the collected program running resource usage into different resource components; for data contained in each resource component, determining a prediction function for the resource component; determining an overall prediction function for a program running resource according to the determined prediction functions for all the resource components; and predicting a usage condition of the program running resource based on the determined overall prediction function.

First claim

What is claimed is:

1. A method for predicting usage of a program resource during running of a program, the method comprising:
collecting, at least once within each period of a plurality of periods of usage of a program resource, resource usage data for the program resource while a program using the program resource is running;
decomposing the resource usage data collected for the plurality of periods into a trend component reflecting a variation trend of program running resource usage, a seasonal component reflecting a periodic variation in the program running resource usage, and a random component reflecting a random variation in the program running resource usage;
determining a first prediction function for the trend component, a second prediction function for the seasonal component, and a third prediction function for the random component, wherein the second prediction function for the seasonal component is an i th data point of the seasonal component, wherein i=t mod T, wherein t is prediction time, and T represents a duration of each period of the plurality of periods;
adding together the first prediction function, the second prediction function, and the third prediction function to obtain an overall prediction function for the program resource; and
predicting the usage of the program resource based on the overall prediction function, wherein predicting the usage includes at least one of predicting program running resource usage at a future set time, predicting when the program resource will be exhausted, or predicting when program running resource usage will reach a set threshold in the future.

2. The method according to claim 1 , wherein the first prediction function for the trend component is a linear function or a nonlinear function that uses prediction time as an independent variable and is obtained by performing linear fitting or nonlinear fitting on resource usage data contained in the trend component, and wherein the third prediction function for the random component is a constant, and the constant is an upper quantile of the random component.

3. The method according to claim 1 , wherein the overall prediction function for the program resource is determined according to the following formula:
R t =( a+bt )+ S i +(μ+ k σ),

wherein in the formula, R t is the overall prediction function for the program resource, (a+bt) is the first prediction function for the trend component, wherein a and b are constants, wherein S i is the second prediction function for the seasonal component, and (μ+kσ) is the third prediction function for the random component, and wherein μ is a mean value of data contained in the random component, σ is a standard deviation of data contained in the random component, k is a constant, and a range of k is (0, 6].

4. The method according to claim 1 , further comprising performing the decomposing based on determining that a central processing unit occupancy rate is less than a second set threshold.

5. The method according to claim 1 , wherein a number of times of collecting resource usage data is the same within each period of the plurality of periods.

6. The method according to claim 1 , further comprising performing the decomposing based on determining that an occupancy rate of the program resource is not less than a second set threshold.

7. An apparatus for predicting usage of a program resource during running of a program, the apparatus comprising:
a computer processor configured to:
collect, at least once within each period of a plurality of periods of usage of a program resource, resource usage data for the program resource while a program using the program resource is running;
decompose the resource usage data collected for the plurality of periods into a trend component reflecting a variation trend of program running resource usage, a seasonal component reflecting a periodic variation in the program running resource usage, and a random component reflecting a random variation in the program running resource usage;
determine a first prediction function for the trend component, a second prediction function for the seasonal component, and a third prediction function for the random component, wherein the second prediction function is an i th data point of the seasonal component, wherein i=t mod T, wherein t is prediction time, and T represents a duration of each period of the plurality of periods;
add together the first prediction function, the second prediction function, and the third prediction function to obtain an overall prediction function for the program resource; and
predict the usage of the program resource based on the overall prediction function, wherein predicting the usage includes at least one of predicting program resource usage at a future set time, predicting when the program resource will be exhausted, or predicting when program resource usage will reach a set threshold in the future.

8. The apparatus according to claim 7 , wherein the first prediction function for the trend component is a linear function or a nonlinear function that uses prediction time as an independent variable and is obtained by performing linear fitting or nonlinear fitting on resource usage data contained in the trend component, and wherein the third prediction function for the random component is a constant, and the constant is an upper quantile of the random component.

9. The apparatus according to claim 7 , wherein the computer processor is configured to determine the overall prediction function for the program resource according to the following formula:
R t =( a+bt )+ S i +(μ+ k σ),

wherein in the formula, R t is the overall prediction function for the program resource, (a+bt) is the first prediction function for the trend component, wherein a and b are constants; wherein S t is the second prediction function for the seasonal component, and (μ+kσ) is the third prediction function for the random component, and wherein μ is a mean value of data contained in the random component, σ is a standard deviation of data contained in the random component, k is a constant, and a range of k is (0, 6].

10. The apparatus according to claim 7 , wherein the computer processor is further configured to initiate the decomposition based on determining that a central processing unit occupancy rate is less than a second set threshold.

11. The apparatus according to claim 7 , wherein a number of times of collecting resource usage data is the same within each period of the plurality of periods.

12. The apparatus according to claim 7 , wherein the computer processor is further configured to initiate the decomposition based on determining that an occupancy rate of the program resource is not less than a second set threshold.

13. A method for determining a leak of a program resource, the method comprising:
collecting, at least once within each period of a plurality of periods of usage of a program resource, resource usage data for the program resource while a program using the program resource is running, wherein a number of times of collecting resource usage data is the same within each period of the plurality of periods;
determining, for every two periods of the plurality of periods and for each program running resource usage sample collected within a latter period, a difference value between the program running resource usage sample collected within the latter period and a corresponding program running resource usage sample collected within a former period, wherein a first time difference between time when collection is performed each time within the latter period and start time of the latter period and a second time difference

Assignees

• Huawei Tech Co Ltd

Inventors

• Li Jinghui
• Gong Xuewen
• Yuan Jianqing

Classifications

• G06F11/3037

  where the computing system component is a memory, e.g. virtual memory, cache (accessing, addressing or allocating within memory systems or architectures G06F12/00; checking stores for correct operation G11C29/00) · CPC title

• G06F9/5055 Primary

  considering software capabilities, i.e. software resources associated or available to the machine · CPC title

• G06F11/3409

  for performance assessment · CPC title

• G06F2201/81

  Threshold · CPC title

• G06F11/008 Primary

  Reliability or availability analysis · CPC title