JP5500301B2 - Monitoring control system, monitoring control method, monitoring control server, and monitoring control program - Google Patents

Description

  The present invention relates to a monitoring control system, a monitoring control method, a monitoring control server, and a monitoring control program for monitoring and controlling a scale-out type server system.

Scale-out type server systems are widely used in relatively large systems such as enterprise information systems, data centers, and providers.
The scale-out type server system is to increase the performance of the entire server group by increasing the number of servers (scale-out).
For example, when one server can process only 10 users, the load is distributed by increasing the number of servers to two, and it becomes possible to deal with 20 users.

In such a scale-out type server system, a plurality of servers (including physical servers and virtual servers) are grouped for each role, and processing requests are equally allocated to the servers in the group. There is an advantage that it can be managed by the number of units.
Even when a server is scaled out, a plurality of servers are operated in a coordinated manner, so there is an advantage that it is not necessary to completely stop the service even when maintenance or a failure occurs.

The spread of such scale-out server systems is also affected by the advent of server provisioning tools that facilitate system configuration changes (scale-outs). The system can be flexibly changed according to the load.
In addition, a capacity planning technique for detecting a bottleneck for the number of transactions per second that can be processed by the system is also provided. According to this capacity planning technology, the system administrator can identify the bottleneck part of the system, estimate the number of servers required for desired performance, and change the system configuration.

JP 2005-309748 A

However, in the scale-out type server system, even if the bottleneck of the current system is identified and the configuration is changed to eliminate the bottleneck, the performance may not be improved.
The reason is that when the allowable load increases with the configuration change, a bottleneck may occur secondarily in another place due to the influence.
And since it is impossible to predict the occurrence of such a secondary bottleneck, after actually changing the configuration, it is necessary to detect the presence or absence of the bottleneck again and examine the configuration change. .

Here, in Patent Document 1, before actually changing the configuration of the storage system, the occurrence of a bottleneck is detected based on the state of each element of the storage system, and a countermeasure for eliminating this bottleneck is presented. A storage system to be used and a failure solving method thereof are shown.
However, the bottleneck elimination method described in Patent Document 1 is specialized in detecting a bottleneck for each element existing on the communication path from the host computer to the storage apparatus. A server that becomes a secondary bottleneck in the server system could not be predicted.

  An object of the present invention is to provide a monitoring control system, a monitoring control method, a monitoring control server, and monitoring control that solve the problem of enabling appropriate configuration change while avoiding the occurrence of the secondary bottleneck, which is the problem described above. To provide a program.

  In order to achieve the above object, a monitoring control system of the present invention includes a monitoring server connected to a plurality of servers constituting the system and monitoring the plurality of servers, and the monitoring server is based on performance information collected from the plurality of servers. A performance modeling means for modeling the performance of each server, and a post-change performance model generating means for generating a performance model after the configuration change from the performance model before the configuration change according to a system configuration change instruction. It is as a structure to provide.

  In addition, the monitoring control method of the present invention connects a monitoring server that monitors a plurality of servers to a plurality of servers that constitute the system, and the monitoring server uses the performance information collected from the plurality of servers to This is a method for modeling performance and generating a performance model after the configuration change from the performance model before the configuration change in accordance with a system configuration change instruction.

  The monitoring control server of the present invention is connected to a plurality of servers constituting the system, and models the performance of each server based on performance information collected from the plurality of servers when monitoring and controlling the plurality of servers. The configuration includes performance modeling means and configuration-changed performance model generation means for generating a performance model after the configuration change from the performance model before the configuration change according to a system configuration change instruction.

  The monitoring control program of the present invention models the performance of each server based on the performance information collected from the plurality of servers in a computer connected to the plurality of servers constituting the system and monitoring and controlling the plurality of servers. In accordance with the system configuration change instruction, the performance model after the configuration change is generated from the performance model before the configuration change.

  According to the present invention, the occurrence of a secondary bottleneck associated with a configuration change of a scale-out type server system is predicted in advance, and an appropriate configuration change is performed while avoiding the occurrence of a secondary bottleneck. can do.

It is a functional block diagram which shows schematic structure of the monitoring control server which concerns on one Embodiment of this invention. It is a functional block diagram which shows the structure of the monitoring control system which concerns on one Embodiment of this invention. It is a flowchart which shows the process sequence of the monitoring control server which concerns on one Embodiment of this invention. It is explanatory drawing which shows the performance item which the monitoring control server which concerns on one Embodiment of this invention collects. In the monitoring control system concerning one embodiment of the present invention, it is a network figure showing correlation established between performance items before server addition. In the monitoring control system which concerns on one Embodiment of this invention, it is a table | surface figure which shows the correlation established between the performance items before a server addition. It is explanatory drawing of the screen which performs the change instruction | indication of a logical structure in the monitoring control system which concerns on one Embodiment of this invention. In the monitoring control system which concerns on one Embodiment of this invention, it is a network diagram which shows the correlation established between the performance items after a server addition. In the monitoring control system which concerns on one Embodiment of this invention, it is a table | surface figure which shows the correlation established between the performance items after a server addition. It is explanatory drawing which shows the allocation of the workload for every server after the server addition in the monitoring control system which concerns on one Embodiment of this invention. It is explanatory drawing of the screen which inputs a desired workload in the monitoring control system which concerns on one Embodiment of this invention. It is a table | surface figure which shows the example of a calculation of the performance value after a server addition in the monitoring control system which concerns on one Embodiment of this invention. It is a table | surface figure which shows the bottleneck detection example after a server addition in the monitoring control system which concerns on one Embodiment of this invention. In the monitoring control system concerning one embodiment of the present invention, it is explanatory drawing of the screen which shows a bottleneck detection result.

Hereinafter, embodiments of a monitoring control system, a monitoring control method, a monitoring control server, and a monitoring control program according to the present invention will be described with reference to the drawings.
Note that the processing operations executed by the monitoring control system and monitoring control server of the present invention described below are realized by processing, means, and functions executed by a computer in accordance with instructions of a program (software). The program sends a command to each component of the computer, and performs predetermined processing of the present invention as described below, for example, modeling of the performance of the server system based on performance information collected from a plurality of monitored control servers, configuration change Generation of performance model after configuration change from previous performance model, prediction of secondary bottleneck occurrence due to server configuration change based on performance model after configuration change, performance information collected from multiple monitored servers Processes / procedures such as modeling of the correlation function of the performance of each monitored control server based on the load information of the server system are performed.

Thus, each process and means in the present invention are realized by specific means in which the program and the computer cooperate.
Note that all or part of the program is provided by, for example, a magnetic disk, optical disk, semiconductor memory, or any other computer-readable recording medium, and the program read from the recording medium is installed in the computer and executed. The The program can also be loaded and executed directly on a computer through a communication line without using a recording medium.

[Overview of monitoring and control system]
First, a schematic configuration of a monitoring control server according to an embodiment of the present invention will be described with reference to FIG.
FIG. 1 is a functional block diagram showing a schematic configuration of a monitoring control server according to an embodiment of the present invention.
As shown in the figure, the monitoring control server 2 of this embodiment is connected to a plurality of monitored control servers 3 constituting a scale-out type server system so as to be communicable, and monitors and controls the plurality of monitored control servers 3. It is a monitoring control server for doing this.

The monitoring control server 2 is composed of, for example, a computer (information processing apparatus) such as a workstation or a personal computer constituting the server, and constitutes the following means by executing a program installed in the computer. It is like that.
First, as shown to Fig.1 (a), the monitoring control server 2 is provided with the bottleneck prediction means 2a.
As shown in FIG. 1B, the monitoring control server 2 can include a performance modeling unit 2b and a post-configuration-change performance model generation unit 2c in addition to the bottleneck prediction unit 2a.

The bottleneck predicting means 2a is a means for predicting the occurrence of a secondary bottleneck accompanying a configuration change of a scale-out type server system composed of a plurality of monitored control servers 3.
The performance modeling unit 2 b is a unit that models the performance of the scale-out type server system configured by the monitored control server 3 based on the performance information collected from the plurality of monitored control servers 3.
The post-configuration-change performance model generation unit 2c is a unit that generates a post-configuration-change performance model from the pre-configuration-change performance model in response to the configuration change instruction of the scale-out server system configured by the monitored control server 3. is there.
The bottleneck predicting means 2a accompanies the configuration change of the scale-out type server system configured by the monitored control server 3 based on the performance model after the configuration change generated by the performance model generating means 2c after the configuration change. The occurrence of secondary bottlenecks is predicted.

The performance modeling unit 2b performs the performance of each monitored control server 3 based on the performance information collected from the plurality of monitored control servers 3 and the load information of the scale-out server system composed of the monitored control servers 3. Is processed into a correlation function with the load as a variable.
The post-configuration-change performance model generation means 2c selects a monitored control server of the same type as the monitored control server 3 to be added from the existing monitored control servers 3 according to the configuration change instruction of the scale-out type server system, A process of replicating the correlation function of the existing monitored control server 3 as the correlation function of the monitored control server to be added is performed.

Further, the bottleneck prediction means 2a assigns a predetermined load to the correlation function of each monitored control server 3 that is the performance model after the configuration change, and determines the monitored control server 3 whose output exceeds a predetermined threshold. Then, it is extracted and predicted as a secondary bottleneck accompanying the configuration change of the scale-out type server system.
Here, the load data to be substituted into the correlation function can be arbitrarily designated / input.
More specific configurations and operations of the monitoring control server 2 according to the present embodiment including the above-described units and the monitoring control system including the same will be described in detail below with reference to FIGS.

[Details of supervisory control system]
FIG. 2 is a block diagram showing the configuration of the monitoring control system according to the embodiment of the present invention.
The system shown in FIG. 1 is a monitoring control system that performs monitoring control of a scale-out type server system, and includes a monitoring terminal 1, a monitoring control server 2, and a plurality of monitored control servers 3.
The monitoring terminal 1 includes an information processing apparatus that is communicably connected to the monitoring control server 2 and includes a performance information display unit 11 and a logical configuration editor 12 as shown in FIG.

The monitoring control server 2 is composed of the same information processing apparatus as the monitoring control server 2 shown in FIG. 1. As shown in FIG. 2, the post-configuration-change performance prediction means 21, the correlation model storage means 23, and the correlation detection unit 24, logical configuration data 25, logical configuration management unit 26, monitoring device 27, performance information storage means 28, and control device 29.
The monitored control server 3 is composed of information processing devices constituting a plurality of server devices communicably connected to the monitoring control server 2, and each monitored control server 3 has a monitoring control agent 31 as shown in FIG. I have.

The monitoring control agent 31 of the monitored control server 3 can communicate with the monitoring device 27 and the control device 29 of the monitoring control server 2 to notify and control performance information of the monitored control server 3.
Here, the performance information of the monitored control server 3 depends on the usage status of the server, but in the present embodiment, since it is a scale-out type server system, transactions are distributed almost evenly among servers of the same type. It is like that.

Examples of the servers constituting the monitored control server 3 include a WEB server, an AP (application) server, a DB (database) server, and the like. Among these servers, servers of the same type (WEB layer, AP layer, DB layer) Transactions are distributed almost evenly between them.
The server performance information is numerical data indicating the usage status of the CPU, memory, network, etc. of the monitored control server 3. These numerical data are collected as time-series data at a predetermined timing (for example, every one minute) for each performance item, and are periodically transmitted to the monitoring device 27.

The time-series performance information collected from the monitored control server 3 is stored in the performance information storage means 28 of the monitoring control server 2.
The correlation detection unit 24 of the monitoring control server 2 extracts a correlation from the collected performance information using a correlation function derivation method described later, and stores it in the correlation model storage unit 23 as a correlation model.

The logical configuration editor 12 of the monitoring terminal 1 receives a logical configuration change request from the user, and instructs the logical configuration management unit 26 of the monitoring control server 2 to change the configuration.
For example, a physical server is added to an arbitrary server group, and the number of active servers is increased to improve the processing performance of the entire system.

The logical configuration management unit 26 of the monitoring control server 2 issues a configuration change instruction to the control device 29 and updates the logical configuration data 25.
The monitored control server 3 is installed with an OS and an application and is set according to the business, but these are changed according to a logical configuration change instruction transmitted via the control device 29.
For example, when the server is diverted to a different use, a new application introduction instruction or an application setting change instruction is sent from the control device 29 of the monitoring control server 2 to the monitoring control agent of the monitored control server 3 together with the application installation file. When the monitoring control agent 31 receives and processes the data, the configuration of the monitored control server 3 is changed.
As a result, a new server can be added to an existing operating system.

When the user inputs / instructs a change in the logical configuration at the monitoring terminal 1, the supervisory control server 2 recognizes the start of the configuration change via the logical configuration data 25 by the performance prediction means 21 after the configuration change.
The post-configuration change performance predicting means 21 calculates a post-configuration change correlation model from the correlation model obtained from the correlation model accumulating means 23, and performs performance prediction after the configuration change.

In the monitoring terminal 1, a throughput value expected after the configuration change is further input to a dialog displayed on the performance information display unit 11 in accordance with a user input operation or the like.
The performance information display unit 11 displays the expected resource consumption required for each server after the configuration change.
If a new bottleneck is detected from this information, a predetermined warning is output and displayed. Thereby, the user who operates the monitoring terminal 1 can recognize the occurrence of the bottleneck before the configuration change, and finally can appropriately change the configuration.

[Processing contents of the monitoring control server]
Next, specific processing contents of the monitoring control server will be described with reference to FIGS.
FIG. 3 is a flowchart showing the processing procedure of the monitoring control server, FIG. 4 is an explanatory diagram showing performance items collected by the monitoring control server, and FIG. 5 is a network diagram showing the correlation established between the performance items before adding the server. 6 is a table showing the correlation established between the performance items before the server addition, FIG. 7 is an explanatory diagram of a screen for instructing the logical configuration change, and FIG. 8 is established between the performance items after the server addition. FIG. 9 is a table showing the correlation established between the performance items after the server addition, FIG. 10 is an explanatory diagram showing the workload assignment for each server after the server addition, and FIG. FIG. 12 is an explanatory diagram of a screen for inputting a desired workload, FIG. 12 is a table showing an example of calculation of a performance value after adding a server, FIG. 13 is a table showing an example of bottleneck detection after adding a server, and FIG. The It is an explanatory view of a screen showing a Runekku detection result.

"Step S1: Performance modeling (performance modeling means)"
First, the “performance modeling” process shown in step S1 of FIG. 3 is performed. This performance modeling process corresponds to the above-described performance modeling means 2b (see FIG. 1).
Specifically, the monitoring control server 2 stores time-series performance information collected from the monitored control server 3 in the performance information storage unit 28. For example, as shown in FIG. 4, the time is stored in the first column, and the performance information is stored in the second column and thereafter.
The monitoring control server 2 also periodically acquires a load (workload) on the system and stores it in the performance information storage means 28. In the example shown in FIG. 4, the load is stored in the last column. The load acquisition method is arbitrary. For example, in the case of a Web server, the load can be acquired by counting the number of requests per second.

Next, the correlation detection unit 24 analyzes the performance information and calculates a correlation function (conversion function) between the performance information as a performance model.
Here, as shown in FIG. 5, the correlation detection unit 24 can select any two items included in the performance information in all combinations, and derive a correlation function for each combination. In the present embodiment, the performance of each monitored control server 2 is modeled as a correlation function with the load as a variable.
Various methods are known as methods for deriving the correlation function. In order to simplify the description, a method for obtaining an approximated linear function “y = Ax + C” will be described.

In the above method, the load is x, the performance is y, and the coefficient of the approximate linear function “y = Ax + C” is obtained.
For example, the correlation function of the element “AP1.CPU” can be derived as “y = 0.5x + 8”, and the values of the coefficients A and C are stored in the correlation model storage unit 23 as shown in FIG. .
Further, when deriving the correlation function, the weight W is also calculated and stored in the correlation model storage unit 23. This weight W is an error correction coefficient of the correlation function, and can be calculated from the difference between the performance pattern obtained by substituting the load into the derived correlation function and the actually collected performance pattern.

"Step S2: Configuration change instruction"
Next, the “configuration change instruction” process shown in step S2 of FIG. 3 is performed.
Specifically, a configuration change instruction is input / designated on the screen of the logical configuration editor 12 of the monitoring terminal 1 (see FIG. 7) by an input operation of the monitoring terminal 1 by a user or the like.
For example, a new AP server named AP2 is added to the AP server group, and the configuration change (server addition) of the AP server layer is instructed / input.
This instruction is transferred to the logical configuration management unit 26 of the monitoring control server 2 and stored in the logical configuration data 25.
The post-configuration change performance prediction means 21 is notified via the logical configuration data 25 that a configuration change (server addition) will be performed.

"Step S3: Recalculation of correlation (performance model generation means after configuration change)"
Next, the process of “recalculation of correlation” shown in step S3 of FIG. 3 is performed. This correlation recalculation process corresponds to the above-described post-configuration-change performance model generation means 2c (see FIG. 1).
Specifically, the post-configuration change performance prediction means 21 generates a post-configuration change performance model (see FIG. 8) from the pre-configuration change performance model in response to the configuration change instruction in step S2.
For example, as shown in FIGS. 5 and 6, in the performance model before the configuration change, the correlation function “y = Ax + C” is established, and if it is desired to add an AP server, the performance prediction means 21 after the configuration change is Select an existing server from the layer where the server is going to be added, and copy the correlation function “y = Ax + C” that holds for the existing server as if it also holds for the new server, and generate a performance model after adding the server To do.
For example, as shown in FIG. 9, the correlation function values A, C, and W are extracted from the data area of “AP1.CPU”, which is an existing server in the AP server layer, and the data area of “AP2.CPU” is extracted. Duplicated.
FIG. 8 shows a correlation established between performance items after configuration change (server addition).

"Step S4: Search for a new bottleneck (bottleneck prediction means)"
Further, a “new bottleneck search” process shown in step S4 of FIG. 3 is performed. This new bottleneck search process corresponds to the bottleneck prediction means 2a described above (see FIG. 1).
Specifically, the post-configuration change performance predicting means 21 detects and predicts the occurrence of a secondary bottleneck associated with the configuration change based on the performance model after the configuration change.
For example, a bottleneck is detected using a performance model by substituting a required value (for example, a required load in this embodiment) into a variable of a correlation function that is a performance model, and the obtained performance value exceeds a predetermined threshold value. It can be detected based on whether or not.
Hereinafter, a specific bottleneck detection method will be described.

As shown in FIG. 10, the load of the same type of server in the scale-out type server system is uniformly distributed. Therefore, for example, when the number of AP servers increases from one to two, the load on each AP server is reduced by half.
The user can input desired request data such as “300 / sec” as a desired load (workload) on the screen as shown in FIG. 11 via the monitoring terminal 1.
The post-configuration-change performance predicting means 21 detects the bottleneck while increasing the load stepwise with the input desired load as the upper limit.
For example, as shown in FIGS. 12 and 13, the bottleneck is detected while starting from “100 / sec” and gradually increasing by “10 / sec”.

Here, when calculating the performance value of each server for the desired load W ′,
y = A ・ W ′ ・ Server additional coefficient + C
And it is sufficient. (However, correction by weight W is omitted)
The server addition factor is for each group of servers.
It can be obtained by calculating the number before adding the server / the number after adding the server.
According to this method, calculations are performed for all performance items.
For example, when calculating the performance value of an AP server,
y = A · W '/ 2 + C
And it is sufficient.
Similarly, when calculating the performance values of the WEB server and DB server,
y = A ・ W '+ C
And it is sufficient.

“Step S5: Judgment”
The determination process shown in step S5 of FIG. 3 is performed.
In this processing step, it is determined whether a bottleneck has been detected.
For example, as shown in FIG. 13, when the user sets the load = 300, the load is 220, and the DB server is saturated in performance.
In this case, as shown in FIG. 14, on the result display screen, the server that is the first bottleneck in the system after the server addition is displayed.
In this way, it is possible to confirm the occurrence of a secondary bottleneck that occurs outside the AP server before adding the AP server.

“Step S6: Warning”
Next, the “warning” process shown in step S6 of FIG. 3 is performed.
When the occurrence of a bottleneck is predicted by the process in step S5 described above, a warning can be urged to the user.
This is because, for example, the addition of one type of server alone cannot supplement the desired throughput, and the addition of other types of servers should be considered at the same time.
Specifically, according to the above-described example, it is possible to determine that the AP server group can process a desired load by reducing the load of the AP server by adding the server, It can be understood that the required load cannot be satisfied.
Therefore, in such a case, a predetermined warning process is executed, and a warning is urged to the user who operates the monitoring terminal 1.

"Step S7: Implementation of configuration change"
Further, the “execution of configuration change” process shown in S step S7 of FIG. 3 is performed.
When it is confirmed that there is no bottleneck by the process in step S5 described above, the configuration is actually changed.
Through the above steps, it is possible to perform an appropriate configuration change without causing a secondary bottleneck associated with the configuration change.

  As described above, according to the monitoring control system and the monitoring control server of this embodiment, the monitoring control server 3 connected to the plurality of monitored control servers 3 constituting the scale-out server system is monitored and controlled. The monitoring control server 2 predicts the occurrence of a secondary bottleneck associated with the configuration change of the scale-out type server system, and thus avoids the occurrence of a secondary bottleneck. Appropriate configuration changes can be made.

  The monitoring control server 2 models the performance of the scale-out type server system based on the performance information collected from the plurality of monitored control servers 3, and before the configuration change according to the configuration change instruction of the scale-out type server system. A performance model after the configuration change is generated from the performance model, and the occurrence of a secondary bottleneck associated with the configuration change of the scale-out server system is predicted based on the performance model after the configuration change. The occurrence of secondary bottlenecks can be accurately predicted based on the performance model before the configuration change.

Further, the monitoring control server 2 correlates the performance of each monitored control server 2 with the load as a variable based on the performance information collected from the plurality of monitored control servers 3 and the load information of the scale-out type server system. Since it is modeled as a function, the performance of each monitored control server 2 with respect to the required load can be easily calculated and a bottleneck can be detected.
The monitoring control server 2 selects the monitored control server 3 of the same type as the monitored control server 3 to be added from the existing monitored control server 3 in response to the configuration change instruction of the scale-out type server system, Since the correlation function of the existing monitored control server 3 is replicated as the correlation function of the monitored control server 3 to be added, the performance model after the configuration change can be easily generated.

Further, the monitoring control server 2 assigns a predetermined load to the correlation function of each monitored control server 3 that is the performance model after the configuration change, and the monitored control server 3 whose output exceeds a predetermined threshold Since it is predicted that this is a secondary bottleneck accompanying the configuration change of the scale-out type server system, it is possible to detect the bottleneck with a simple calculation process.
Furthermore, when predicting a secondary bottleneck associated with a configuration change in a scale-out server system, the load to be assigned to the correlation function is arbitrarily specified, so the user can change the configuration by inputting various loads. The accompanying secondary bottleneck can be predicted and appropriate configuration changes can be made.

The present invention has been described with reference to the embodiment. However, the present invention is not limited to the above-described embodiment, and various modifications can be made within the scope of the claims. .
For example, in the above-described embodiment, three types of servers, a WEB server, an AP (application) server, and a DB (database) server, have been described as examples of servers provided in the monitored control server. The type and number of the target servers are not particularly limited to the above-described embodiment.

  The present invention can be used as a monitoring control system, a monitoring control method, a monitoring control server, and a monitoring control program for monitoring and controlling a scale-out type server system.

DESCRIPTION OF SYMBOLS 1 Monitoring terminal 11 Performance information display part 12 Logical configuration editor 2 Monitoring control server 21 Performance prediction means 23 after a configuration change Correlation model storage means 24 Correlation detection part 25 Logical configuration data 26 Logical configuration management part 27 Monitoring apparatus 28 Performance information storage means 29 Control device 3 Monitored control server 31 Monitoring control agent

Claims (9)

A monitoring server connected to a plurality of servers constituting the system and monitoring the plurality of servers;
The monitoring server is
A performance modeling means for modeling the performance of each server based on performance information collected from a plurality of the servers;
A monitoring control system, comprising: a post-configuration-change performance model generation unit that generates a post-configuration-change performance model from a pre-configuration-change performance model in response to the system configuration change instruction. The monitoring server is
The monitoring control system according to claim 1, further comprising: a bottleneck prediction unit that predicts the occurrence of a secondary bottleneck associated with the system configuration change based on the performance model after the configuration change. The performance modeling means comprises:
The monitoring control system according to claim 2, wherein the performance of each server is modeled as a correlation function having a load as a variable based on performance information collected from a plurality of servers and load information of the system. . The performance model generation means after the configuration change,
According to the system configuration change instruction, a server of the same type as the server to be added is selected from existing servers, and the correlation function of the existing server is copied as the correlation function of the added server. The monitoring control system according to claim 3. The bottleneck prediction means is
A server in which a predetermined load is assigned to the correlation function of each server, which is a performance model after the configuration change, and the output exceeds a predetermined threshold is a secondary bottleneck accompanying the configuration change of the system The monitoring control system according to claim 4, wherein prediction is performed. The monitoring control system according to claim 5, wherein when predicting a secondary bottleneck associated with the system configuration change, a load to be substituted into a correlation function is arbitrarily designated. Connecting a plurality of monitoring servers that monitor the plurality of servers to a plurality of servers constituting the system,
The monitoring server is
Based on performance information collected from multiple servers, model the performance of each server,
A monitoring control method, comprising: generating a performance model after a configuration change from a performance model before the configuration change according to a configuration change instruction of the system. In monitoring and controlling a plurality of servers connected to a plurality of servers constituting the system,
A performance modeling means for modeling the performance of each server based on performance information collected from a plurality of the servers;
A monitoring control server, comprising: a configuration model after configuration change that generates a performance model after the configuration change from the performance model before the configuration change in response to the configuration change instruction of the system. A computer connected to a plurality of servers constituting the system and monitoring and controlling the plurality of servers;
Based on performance information collected from multiple servers, model the performance of each server,
A monitoring control program for generating a performance model after a configuration change from a performance model before the configuration change in response to a configuration change instruction of the system.

Images