KR-102962700-B1 - METHOD, COMPUTER DEVICE, AND COMPUTER PROGRAM FOR PREDICTING AND UTILIZING FUTURE SYSTEM USAGE

KR102962700B1KR 102962700 B1KR102962700 B1KR 102962700B1KR-102962700-B1

Abstract

A method, a computer device, and a computer program for accurate prediction and utilization of future system usage are disclosed. The method for predicting future system usage may include: a step of constructing a prediction model for an input indicator using time series data of an input indicator stored during a first predetermined period; a step of obtaining a correlation between the input indicator and the system performance indicator using time series data of the input indicator and the system performance indicator stored during a second predetermined period; a step of predicting an input indicator at a future time point through the prediction model for at least one future time point; and a step of calculating a system performance indicator at a future time point using the predicted input indicator and the correlation.

Inventors

김완호
최철규
곽효영
류재훈
양은숙
정재부

Assignees

네이버 주식회사

Dates

Publication Date: 20260511
Application Date: 20231227

Claims (20)

A method for predicting future system usage of a computer device including at least one processor, A step of constructing a prediction model for an input indicator using time series data of an input indicator stored during a first specified period by the above-mentioned at least one processor; A step of determining the correlation between the input indicator and the system performance indicator using time series data of the input indicator and the system performance indicator stored during a second predetermined period by the at least one processor; The step of predicting an input indicator of a future point in time through the prediction model for at least one future point in time by the at least one processor; and A step of calculating a system performance indicator at a future point in time using the predicted input indicator and the correlation by the above at least one processor Includes, The step of building the above prediction model is, A step of analyzing input and output indicators during the first specified period and selecting an indicator whose change during the third specified period is less than a threshold as an indicator to be used as training data for the prediction model. including A method for predicting future system usage characterized by
delete
In paragraph 1, User traffic corresponding to log data entering from the outside is used as the above input metric, and CPU usage or disk usage is used as the above system performance indicator. A method for predicting future system usage characterized by
In paragraph 1, The step of calculating the above correlation is, Calculating the correlation between the above input indicators and the above system performance indicators through a linear regression model A method for predicting future system usage characterized by
In paragraph 1, The step of calculating the above correlation is, A step of determining linear regression coefficients and linear regression intercepts through linear regression on the time series data of the second specified period. A method for predicting future system usage including
In paragraph 5, The above linear regression intercept value is determined to be a value close to 0. A method for predicting future system usage characterized by
In paragraph 1, The step of building the above prediction model is, A step of updating the prediction model using data stored since the previous prediction point for the above input indicators. A method for predicting future system usage including
In paragraph 1, The step of calculating the above correlation is, A step of updating the correlation between the above input indicator and the above system performance indicator at regular intervals A method for predicting future system usage including
In paragraph 1, The step of calculating the above correlation is, Depending on the type of the above system performance indicator, for some types, the correlation between the above input indicator and the above system performance indicator is calculated using the value obtained by integrating the time series data of the above input indicator over time. A method for predicting future system usage characterized by
In paragraph 1, The above method for predicting future system usage is, A step of recommending a component configuration at a future point in time using the calculated system performance indicators by the above at least one processor A method for predicting future system usage that further includes
In Paragraph 10, The step of calculating system performance indicators at the aforementioned future point in time is, Calculate system performance indicators at the aforementioned future point in time for each component constituting the cluster, and The step of recommending the component configuration at the aforementioned future point in time is, Calculating and recommending the number of equipment units for each of the above components based on the system performance indicators at the above future point in time. A method for predicting future system usage characterized by
In paragraph 1, The above method for predicting future system usage is, A step of calculating system performance indicators for each component constituting the new cluster using the correlation between the input indicators scheduled to be received and the input indicators scheduled to be received, when the input indicators scheduled to be received are provided by the at least one processor; and A step of calculating and recommending the number of required equipment units for each component based on system performance indicators for each component by the above-mentioned at least one processor. A method for predicting future system usage that further includes
In paragraph 1, The above method for predicting future system usage is, The step of providing, by the at least one processor, a point in time among the future points in time when the calculated system performance indicator deviates from a critical range. A method for predicting future system usage that further includes
A computer program stored on a non-transient computer-readable recording medium to execute the method for predicting future system usage of claim 1, or any one of claims 3 through 13, on the computer device.
At least one processor implemented to execute readable instructions on a computer device Includes, By the above at least one processor, A prediction model for the input indicator is constructed using time series data of the input indicator stored during a first specified period, and Using time series data of the input indicator and system performance indicator stored during a second specified period, the correlation between the input indicator and the system performance indicator is calculated, and For at least one future point in time, the input indicator of the future point in time is predicted through the prediction model, and Calculate the system performance indicators at the future point in time using the above predicted input indicators and the above correlation, and To construct the above prediction model, by the above at least one processor, Analyzing the input and output indicators during the first specified period and selecting the indicator whose change during the third specified period is less than a threshold as the indicator to be used as training data for the prediction model. A computer device characterized by
delete
In paragraph 15, By the above at least one processor, The prediction model is updated using data stored since the previous prediction point for the above input indicators, and Updating the correlation between the above input indicators and the above system performance indicators at regular intervals A computer device characterized by
In paragraph 15, By the above at least one processor, Depending on the type of the above system performance indicator, for some types, the correlation between the above input indicator and the above system performance indicator is calculated using the value obtained by integrating the time series data of the above input indicator over time. A computer device characterized by
In paragraph 15, By the above at least one processor, Calculate system performance indicators at the aforementioned future point in time for each component constituting the cluster, and Calculating and recommending the number of equipment units for each of the above components based on the system performance indicators at the above future point in time. A computer device characterized by
In paragraph 15, By the above at least one processor, Providing a point in time among the above future points in time when the above calculated system performance indicators exceed the critical range. A computer device characterized by

Description

Method, computer device, and computer program for accurate prediction and utilization of future system usage The following description concerns technology for predicting system usage. A cluster is a system used to connect independent computer systems, such as personal computers, using network equipment for specific purposes, and is widely used today as a solution to provide high-performance computing and high-availability computing. It offers the advantage of flexible scalability, as only specific nodes or networks can be replaced or added without overhauling the entire system when upgrades or expansions are required. Due to this advantage, organizations such as companies, research institutes, and university laboratories that need to execute complex scientific calculations build and use clusters tailored to their specific computational demands. For such clusters, monitoring technology is required to continuously observe and maintain the normal operation and quality of the cluster. For example, Korean Patent Publication No. 10-2009-0061522 (publication date June 16, 2009) discloses a large-scale cluster monitoring technology that can always increase availability by actively responding to configuration information in the event of a node failure in a cluster system environment. FIG. 1 is a drawing illustrating an example of a network environment according to an embodiment of the present invention. FIG. 2 is a block diagram illustrating an example of a computer device according to an embodiment of the present invention. FIG. 3 illustrates a system example of a cluster environment in one embodiment of the present invention. FIG. 4 is a flowchart illustrating an example of a method that a computer device according to an embodiment of the present invention can perform. FIGS. 5 and 6 illustrate examples of input indicators and system performance indicators in an embodiment of the present invention. FIG. 7 is a flowchart illustrating the CPU usage prediction process in one embodiment of the present invention. FIG. 8 illustrates an example of the predicted result of CPU usage in one embodiment of the present invention. FIG. 9 illustrates an example of a configuration file for predicting resource usage in an embodiment of the present invention. FIGS. 10 to 12 illustrate examples of component configuration recommendations based on resource usage prediction in an embodiment of the present invention. Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings. Embodiments of the present invention relate to a technique for predicting system usage. Embodiments including those specifically disclosed in this specification can provide an accurate prediction of future system usage by calculating future system performance indicators through the prediction of changes in input indicators, by training a time-series prediction model with an input indicator having a low number of sudden change factors as an input indicator correlated with system performance indicators, and constructing a prediction model for predicting changes in input indicators. A future system usage prediction device according to embodiments of the present invention may be implemented by at least one computer device, and a future system usage prediction method according to embodiments of the present invention may be performed through at least one computer device included in the future system usage prediction device. At this time, a computer program according to an embodiment of the present invention may be installed and run on the computer device, and the computer device may perform a future system usage prediction method according to embodiments of the present invention under the control of the run computer program. The above-described computer program may be stored on a computer-readable recording medium to be combined with the computer device to execute the future system usage prediction method on the computer. FIG. 1 is a diagram illustrating an example of a network environment according to an embodiment of the present invention. The network environment of FIG. 1 illustrates an example including a plurality of electronic devices (110, 120, 130, 140), a plurality of servers (150, 160), and a network (170). FIG. 1 is an example for explaining the invention, and the number of electronic devices or servers is not limited to that shown in FIG. 1. Furthermore, the network environment of FIG. 1 is merely an example of one of the environments applicable to the present embodiments, and the environments applicable to the present embodiments are not limited to the network environment of FIG. 1. Multiple electronic devices (110, 120, 130, 140) may be fixed terminals or mobile terminals implemented as computer devices. Examples of multiple electronic devices (110, 120, 130, 140) include smartphones, mobile phones, navigation systems, computers, laptops, digital broadcasting terminals, PDAs (Personal Digital Assista