Analysis of time-series data indicating temporal variation in usage states of resources used by multiple processes

Patent metadata

Field	Value
Publication number	US-11550770-B2
Application number	US-201916579148-A
Country	US
Kind code	B2
Filing date	Sep 23, 2019
Priority date	Oct 4, 2018
Publication date	Jan 10, 2023
Grant date	Jan 10, 2023

Abstract

Time-series data indicating a temporal variation of an index, which indicates a usage state of each of resources that are used by multiple processes, is acquired, and an operation-data matrix including vectors is generated based on the time-series data such that each of the vectors indicates the time-series data at a predetermined time interval and includes as an element the index indicating the usage state of one of the resources at the predetermined time interval. A basis matrix including a predetermined number of basis vectors is generated by performing nonnegative matrix factorization on the operation-data matrix. Component values, which respectively correspond to the resources, indicated by each of the predetermined number of the basis vectors are extracted, and information on the extracted component values is output as usage states of the resources that are used by each of the multiple processes.

First claim

What is claimed is:

1. A non-transitory, computer-readable recording medium having stored therein a program for causing a computer to execute a process comprising:
acquiring time-series data indicating a temporal variation of an index, the index indicating a usage state of each of one or more resources that are used by multiple processes;
forming a load model in which the index is proportional to a processing amount of each of the multiple processes, the load model to separate, from the time-series data associated with each of the one or more resources, pieces of component data having corresponding periodical variation tendencies at predetermined time intervals;
generating, based on the acquired time-series data, an operation-data matrix including vectors as columns or rows such that each of the vectors indicates the time-series data at a predetermined time interval and includes as an element the index indicating the usage state of one of the one or more resources at the predetermined time interval;
performing nonnegative matrix factorization on the generated operation-data matrix to generate a basis matrix including a predetermined number of basis vectors as columns or rows, the basis vectors indicating the component data of the one or more resources having corresponding periodical variation tendencies at the predetermined time intervals;
extracting one or more component values indicated by each of the predetermined number of the basis vectors included in the generated basis matrix, the extracted one or more components values corresponding to the one or more resources, respectively; and
outputting information on the extracted one or more component values as usage states of the one or more resources that are used by each of the multiple processes.

2. The non-transitory, computer-readable recording medium of claim 1 , the process further comprising:
performing nonnegative matrix factorization on the operation-data matrix to generate a weight matrix including the predetermined number of weight vectors as rows or columns such that the generated matrix is represented as a product of the basis matrix and the generated weight matrix; and
outputting information on each of the predetermined number of the weight vectors included in the weight matrix.

3. The non-transitory, computer-readable recording medium of claim 1 , wherein
the predetermined number is set based on a number of processes to be performed by using at least one of the one or more resources.

4. The non-transitory, computer-readable recording medium of claim 1 , wherein
the predetermined number is set based on a number of dimensions of the vectors.

5. The non-transitory, computer-readable recording medium of claim 1 , wherein
the one or more resources include different types of resources in a single device.

6. The non-transitory, computer-readable recording medium of claim 1 , wherein
the one or more resources include resources in different devices.

7. The non-transitory, computer-readable recording medium of claim 1 , the process further comprising:
dividing the time-series data into pieces of the time-series data such that the pieces of the time-series data respectively correspond to time intervals each equal to the predetermined time interval; and
generating the operation-data matrix including vectors such that the vectors include as elements one or more indexes indicated by the pieces of the time-series data.

8. A method performed by a computer, the method comprising:
acquiring time-series data indicating a temporal variation of an index, the index indicating a usage state of each of one or more resources that are used by multiple processes;
forming a load model in which the index is proportional to a processing amount of each of the multiple processes, the load model to separate, from the time-series data associated with each of the one or more resources, pieces of component data having corresponding periodical variation tendencies at predetermined time intervals;
generating, based on the acquired time-series data, an operation-data matrix including vectors as columns or rows such that each of the vectors indicates the time-series data at a predetermined time interval and includes as an element the index indicating the usage state of one of the one or more resources at the predetermined time interval;
performing nonnegative matrix factorization on the generated operation-data matrix to generate a basis matrix including a predetermined number of basis vectors as columns or rows, the basis vectors indicating the component data of the one or more resources having corresponding periodical variation tendencies at the predetermined time intervals;
extracting one or more component values indicated by each of the predetermined number of the basis vectors included in the generated basis matrix, the extracted one or more components values corresponding to the one or more resources, respectively; and
outputting information on the extracted one or more component values as usage states of the one or more resources that are used by each of the multiple processes.

9. An apparatus comprising:
a memory; and
a processor coupled to the memory and configured to:
acquire time-series data indicating a temporal variation of an index, the index indicating a usage state of each of one or more resources that are used by multiple processes,
form a load model in which the index is proportional to a processing amount of each of the multiple processes, the load model to separate, from the time-series data associated with each of the one or more resources, pieces of component data having corresponding periodical variation tendencies at predetermined time intervals,
generate, based on the acquired time-series data, an operation-data matrix including vectors as columns or rows such that each of the vectors indicates the time-series data at a predetermined time interval and includes as an element the index indicating the usage state of one of the one or more resources at the predetermined time interval,
perform nonnegative matrix factorization on the generated operation-data matrix to generate a basis matrix including a predetermined number of basis vectors as columns or rows, the basis vectors indicating the component data of the one or more resources having corresponding periodical variation tendencies at the predetermined time intervals,
extract one or more component values indicated by each of the predetermined number of the basis vectors included in the generated basis matrix, the extracted one or more components values corresponding to the one or more resources, respectively, and
output information on the extracted one or more component values as usage states of the one or more resources that are used by each of the multiple processes.

Assignees

• Fujitsu Ltd

Inventors

• Saito Yuji
• Uchiumi Tetsuya
• Watanabe Yukihiro

Classifications

• G06F16/2237 Primary

  Vectors, bitmaps or matrices · CPC title

• G06F16/902

  using more than one table in sequence, i.e. systems with three or more layers · CPC title

• G06F16/211

  Schema design and management · CPC title

• G06F16/2477 Primary

  Temporal data queries · CPC title

• G06F16/258

  Data format conversion from or to a database · CPC title