JP2021101278A - Information processing device, information processing system, information processing method, and program - Google Patents

Abstract

To provide an information processing device which reduces a communication volume for collecting operation information.SOLUTION: An information processing device 10 includes a communication unit 11, a storage unit 12, and a processing unit 13. The communication unit receives operation information 23 indicating a usage state of resources from a virtual machine 21, and receives operation information 24 indicating a usage state of resources from a virtual machine 22. The storage unit stores histories of the received operation information 23 and 24. The processing unit determines a correlation 14 of the resource usage states of the virtual machine 21 and the virtual machine 22, and when the correlation satisfies a prescribed condition, performs control so that a reception frequency of the operation information 24 is lower than a reception frequency of the operation information 23. The processing unit estimates the resource usage state in the virtual machine 22 on the basis of the operation information 23 received from the virtual machine 21 and the correlation.SELECTED DRAWING: Figure 1

Description

The present invention relates to an information processing device, an information processing system, an information processing method and a program.

There is a computer virtualization technology in which a virtual machine, which is a virtual computer, is placed and executed on a physical machine, which is a physical computer. According to computer virtualization technology, virtual machines can be dynamically added, deleted, and moved, facilitating the operation management of information processing. The physical machine in which the virtual machine is placed may be owned by a large-scale information processing facility such as a data center, or may be owned by a so-called cloud system.

Multiple virtual machines located on different physical machines can be distributed to execute application programs, and these multiple virtual machines can work together to provide information processing services in response to user requests. .. It is also possible to divide and arrange multiple virtual machines in different data centers and different cloud systems. In the operation management of an information processing service using a plurality of virtual machines, it is preferable to collect operation information indicating the usage status of resources such as processors and memories from each virtual machine. Based on the collected operation information, it is possible to consider adding new virtual machines, deleting some virtual machines, increasing the resources allocated to virtual machines, and changing the cloud system in which virtual machines are placed.

For example, a virtual machine management device that collects measurement data showing the resource usage rate of the entire physical machine and the resource usage rate of each virtual machine, and moves some virtual machines to other physical machines when a resource shortage is detected. Has been proposed. Further, when moving two virtual machines from one physical machine to another, a virtual machine control device for determining whether or not to move the two virtual machines in parallel has been proposed. When the proposed virtual machine controller starts moving one virtual machine, it monitors the processing status of the two virtual machines and calculates the correlation coefficient. If the correlation coefficient is positive, the other virtual machine The movement of is also started.

International Publication No. 2008/062864 Japanese Unexamined Patent Publication No. 2012-888808

From the viewpoint of the accuracy of the operation information to be collected, it is preferable to collect the operation information of a plurality of virtual machines in a short cycle. However, if communication is performed with all the virtual machines used in the information processing service in a short cycle, there is a problem that the amount of communication when collecting operation information increases. For example, when at least a part of a plurality of virtual machines is arranged in a data center or a cloud system, communication is performed in a short cycle via a wide area data communication network such as the Internet, and the communication load becomes a problem.

In one aspect, it is an object of the present invention to provide an information processing device, an information processing system, an information processing method and a program capable of reducing the amount of communication when collecting operation information.

In one aspect, an information processing device having a communication unit, a storage unit, and a processing unit is provided. The communication unit receives the first operation information indicating the resource usage status in the first virtual machine from the first virtual machine, and indicates the resource usage status in the second virtual machine from the second virtual machine. Receive the operation information of 2. The storage unit stores the received first operation information and the history of the second operation information. The processing unit determines the correlation of the resource usage status between the first virtual machine and the second virtual machine based on the history, and if the correlation satisfies a predetermined condition, the second operation information The reception frequency is controlled to be lower than the reception frequency of the first operation information. The processing unit estimates the resource usage status in the second virtual machine based on the first operation information received from the first virtual machine and the correlation.

Further, in one embodiment, an information processing system having a first information processing device that executes a first virtual machine, a second information processing device that executes a second virtual machine, and a third information processing device. Provided. Also, in one aspect, an information processing method executed by a computer is provided. Also, in one aspect, a program to be executed by a computer is provided.

On one side, the amount of communication when collecting operation information can be reduced.

It is a figure explaining the example of the information processing system of 1st Embodiment. It is a figure which shows the example of the information processing system of the 2nd Embodiment. It is a block diagram which shows the hardware example of an information collection server. It is a figure which shows the correspondence example of a service, a virtual machine, and a system. It is a sequence diagram which shows the communication example of an information collection server and a virtual machine. It is a figure which shows the grouping example of a virtual machine. It is a figure which shows the grouping example of a virtual machine (continuation 1). It is a figure which shows the grouping example of a virtual machine (continuation 2). It is a block diagram which shows the functional example of an information collection server and a virtual machine. It is a figure which shows the example of the service structure table and the operation history table. It is a figure which shows the example of the correlation table and the operation status table. It is a figure which shows the example of the group management table. It is a flowchart which shows the procedure example of a server start processing. It is a flowchart which shows the procedure example of group determination. It is a flowchart which shows the procedure example of the virtual machine start processing. It is a flowchart which shows the procedure example of virtual machine continuation processing. It is a flowchart which shows the procedure example of a server continuation processing.

Hereinafter, the present embodiment will be described with reference to the drawings.
[First Embodiment]
The first embodiment will be described.

FIG. 1 is a diagram illustrating an example of an information processing system according to the first embodiment.
In the information processing system of the first embodiment, a plurality of virtual machines are arranged in different information processing devices, and the resource usage status of each of the plurality of virtual machines is monitored. The information processing system of the first embodiment includes information processing devices 10, 20, 20a.

The information processing device 10 is a computer that monitors a virtual machine by collecting operation information indicating the usage status of resources in the virtual machine. The information processing device 10 may be a client computer or a server computer.

The information processing devices 20 and 20a are server computers capable of executing virtual machines. The information processing devices 20 and 20a can communicate with the information processing device 10 via a network. The information processing devices 20 and 20a may be computers in a data center or computers in a public cloud system owned by a cloud operator. Further, the information processing devices 20 and 20a may be computers of a private cloud system or an on-premises system owned by a service provider that provides services to users by using a virtual machine. The information processing device 20 and the information processing device 20a may belong to different systems. The information processing devices 20 and 20a and the information processing device 10 may communicate with each other via a wide area data communication network such as the Internet.

A virtual machine 21 (first virtual machine) is arranged in the information processing device 20. A virtual machine 22 (second virtual machine) is arranged in the information processing device 20a. The information processing device 10 monitors the resource usage status of each of the virtual machines 21 and 22. For example, the information processing device 10 presents the resource usage status in the virtual machines 21 and 22 to the service manager by a method such as displaying it on the display device. By monitoring resource usage, you can consider configuration changes such as adding new virtual machines, deleting some virtual machines, increasing the resources allocated to virtual machines, and changing the system where virtual machines are located.

A resource is a hardware resource that is allocated to a virtual machine and used for information processing. Resource usage includes processor usage such as CPU (Central Processing Unit), memory usage such as RAM (Random Access Memory), storage access such as HDD (Hard Disk Drive), and network bandwidth usage. Can be mentioned.

The virtual machine 21 can transmit operation information 23 (first operation information) indicating the resource usage status in the virtual machine 21. The virtual machine 22 can transmit operation information 24 (second operation information) indicating the resource usage status in the virtual machine 22. The virtual machines 21 and 22 may transmit operation information in response to a request from the information processing device 10, or autonomously transmit operation information from the information processing device 10 at a frequency specified directly or indirectly. You may. The decrease in transmission frequency, which will be described later, includes stopping the transmission of operation information. The virtual machines 21 and 22 may be virtual machines in a narrow sense that execute a guest operating system (OS), respectively, or may be a virtual information processing entity that does not execute a guest OS such as a container in container-type computer virtualization technology. ..

The information processing device 10 includes a communication unit 11, a storage unit 12, and a processing unit 13. The communication unit 11 is a communication interface that communicates with the information processing devices 20 and 20a via a network. The communication unit 11 may be a wired communication interface connected to a wired communication device such as a switch or a router, or a wireless communication interface connected to a wireless communication device such as an access point or a wireless base station. The storage unit 12 may be a volatile semiconductor memory such as a RAM, or a non-volatile storage such as an HDD or a flash memory.

The processing unit 13 is, for example, a processor such as a CPU, a GPU (Graphics Processing Unit), or a DSP (Digital Signal Processor). However, the processing unit 13 may include an electronic circuit for a specific purpose such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). The processor executes a program stored in a memory (may be a storage unit 12) such as a RAM. A set of multiple processors is sometimes referred to as a "multiprocessor" or simply a "processor."

The communication unit 11 receives the operation information 23 from the virtual machine 21 and the operation information 24 from the virtual machine 22. The storage unit 12 stores the history of the operation information 23 received from the virtual machine 21 and the operation information 24 received from the virtual machine 22. The history of the operation information 23 and 24 shows the usage status of the resource at each of a plurality of different times. Until a sufficient history is accumulated in the storage unit 12, for example, the communication unit 11 receives operation information from each of the virtual machines 21 and 22 at a low predetermined frequency (for example, at 1-minute intervals). The processing unit 13 may notify the virtual machines 21 and 22 of the initial value of the transmission frequency so that the virtual machines 21 and 22 may periodically transmit the operation information. Further, the communication unit 11 may collect the operation information by requesting the virtual machines 21 and 22 for the operation information at a predetermined frequency.

The processing unit 13 determines the correlation 14 of the resource usage status between the virtual machine 21 and the virtual machine 22 based on the history stored in the storage unit 12. When the operation information 23 and 24 indicate the usage status of two or more types of resources, the processing unit 13 may determine the correlation 14 for each type of resource. For example, the processing unit 13 correlates the operation information 23 and the operation information 24 whose times are close to each other from the history stored in the storage unit 12, and based on a plurality of pairs, the numerical value included in the operation information 23 and the operation. The correlation coefficient between the numerical values included in the information 24 is calculated. Further, for example, the processing unit 13 calculates the average magnification of the numerical value included in the operation information 24 with respect to the numerical value included in the operation information 23 based on the operation information 23 and 24 in the latest fixed period.

When the correlation 14 satisfies a predetermined condition, the processing unit 13 controls so that the reception frequency of the operation information 24 of the virtual machine 22 is lower than the reception frequency of the operation information 23 of the virtual machine 21. The predetermined condition is, for example, that the absolute value of the correlation coefficient is larger than the threshold value (for example, 0.8). At this time, the processing unit 13 may set the reception frequency of the operation information 24 to be lower than the initial value before determining the correlation 14, or may stop the reception of the operation information 24 from the virtual machine 22. Further, the processing unit 13 may set the reception frequency of the operation information 23 to be higher than the initial value before determining the correlation 14, such as setting the 1-minute interval to the half of the 30-second interval.

In the control of the reception frequency, the processing unit 13 may notify the virtual machine 21 of the information about the virtual machine 22 to instruct the virtual machine 22 to reduce the transmission frequency of the operation information 24 through the virtual machine 21. .. Further, the processing unit 13 may directly instruct the virtual machine 22 to reduce the transmission frequency of the operation information 24. The instruction given directly or indirectly to the virtual machine 22 may be an instruction to stop transmission. Further, the frequency of requesting the operation information 24 from the communication unit 11 to the virtual machine 22 may be reduced.

The virtual machine whose reception frequency is lowered when the correlation 14 satisfies a predetermined condition may be a virtual machine located on the downstream side of the data flow among the virtual machines 21 and 22. For example, the communication unit 11 transmits a test request and collects the communication logs of the virtual machines 21 and 22, and the processing unit 13 detects the order of message communication between the plurality of virtual machines. When the virtual machine 21 is the source of the message and the virtual machine 22 is the destination of the message, the processing unit 13 determines the reception frequency of the virtual machine 22 located on the downstream side of the data flow among the virtual machines 21 and 22. It is possible to lower it.

When the reception frequency of the operation information 24 becomes lower than the reception frequency of the operation information 23, the operation information 23 indicating the resource usage status of the virtual machine 21 is collected, while the operation information 24 indicating the resource usage status of the virtual machine 22 is collected. There will be periods when it will not be collected. Therefore, the processing unit 13 estimates the resource usage status of the virtual machine 22 during the period when the operation information 24 is not collected. Specifically, the processing unit 13 estimates the resource usage status of the virtual machine 22 based on the operation information 23 received from the virtual machine 21 and the correlation 14. For example, the processing unit 13 estimates that the numerical value included in the operation information 23 is multiplied by the magnification calculated when determining the correlation 14 as a numerical value indicating the resource usage status of the virtual machine 22. ..

The processing unit 13 may mix and present the numerical value included in the operation information 23 and the estimated numerical value to the service manager. For example, the processing unit 13 displays the numerical value included in the latest operation information 23 for the resource usage status of the virtual machine 21, and estimates the resource usage status of the virtual machine 22 based on the correlation 14. Display the numerical value.

According to the information processing apparatus 10 of the first embodiment, the history of the operation information 23 received from the virtual machine 21 and the operation information 24 received from the virtual machine 22 is accumulated. Based on the history, the correlation 14 of the resource usage status between the virtual machine 21 and the virtual machine 22 is determined, and when the correlation 14 satisfies a predetermined condition, the reception frequency of the operation information 24 of the virtual machine 22 is high. It gets lower. Then, the resource usage status of the virtual machine 22 is estimated based on the operation information 23 received from the virtual machine 21 and the correlation 14.

As a result, the frequency of collecting operation information from some virtual machines becomes low, and the amount of communication can be reduced as compared with the case of collecting operation information from all virtual machines at high frequency. In particular, when a plurality of virtual machines are arranged in different systems (for example, different cloud systems), it is possible to reduce the traffic of a wide area data communication network such as the Internet. In addition, the operation information of virtual machines with low collection frequency (including virtual machines for which collection is stopped) is estimated from the correlation with the operation information of virtual machines with high collection frequency. Therefore, it is possible to improve the monitoring accuracy of monitoring the resource usage status as compared with the case where the collection frequency of all virtual machines is uniformly reduced.

In the information processing system of the first embodiment, an estimation process for estimating the resource usage status of the virtual machine 22 is performed, and the addition of the estimation process may slightly increase the load on the information processing system. is there. On the other hand, by adopting the estimation process, the monitoring process of transmitting the operation information 24 from the virtual machine 22 to the information processing device 10 and processing the operation information 24 is reduced, and the reduction of the monitoring process greatly reduces the load on the information processing system. .. Therefore, the effect of reducing the load of the monitoring process is greater than the increase of the load of the estimation process, and the overall load of the information processing system is reduced.

Further, a distributed management is adopted in which the virtual machine 22 whose collection frequency is reduced is subordinated to the virtual machine 21 by transmitting various instructions to the virtual machine 22 through the virtual machine 21, and the monitoring of the virtual machine 22 is entrusted to the virtual machine 21. You can also do it. As a result, communication between the information processing device 10 and the virtual machine 22 can be reduced. For example, when the virtual machines 21 and 22 are arranged in the same cloud system and the information processing device 10 exists outside the cloud system, the communication of the wide area data communication network can be reduced.

[Second Embodiment]
Next, a second embodiment will be described.
FIG. 2 is a diagram showing an example of the information processing system of the second embodiment.

The information processing system of the second embodiment is a multi-cloud system in which a plurality of virtual machines used for providing services are distributed and arranged in different systems. A broad-sense multi-cloud system includes a narrow-sense multi-cloud system and a hybrid cloud system. A multi-cloud system in a narrow sense is a system in which a plurality of virtual machines are distributed and arranged in two or more public cloud systems. In the hybrid cloud system, a plurality of virtual machines are distributed and arranged in a public cloud system and an in-house system (on-premises system). By using multiple systems, it is possible to reduce the risk of service outages and utilize system-specific functions.

The information processing system of the second embodiment includes cloud systems 31, 32, an in-house system 33, and an information collection server 100. The cloud systems 31 and 32, the in-house system 33, and the information collection server 100 are connected to the network 30. The network 30 is a wide area data communication network such as the Internet.

The cloud system 31 is a public cloud system that is owned by a cloud operator and allows the service provider to place virtual machines in accordance with a contract with the cloud operator. The cloud system 31 includes a plurality of servers such as servers 200 and 200a, which are physical machines that execute virtual machines.

The cloud system 32 is a public cloud system that is owned by a cloud operator different from the cloud system 31 and in which the service provider can arrange virtual machines according to a contract with the cloud operator. The cloud system 32 has a plurality of servers such as servers 200b and 200c, which are physical machines that execute virtual machines.

The in-house system 33 is an on-premises system owned by a service provider that provides services using virtual machines. The in-house system 33 includes a plurality of servers such as servers 200d and 200e that execute virtual machines. The service provider distributes and arranges a plurality of virtual machines in two or more of the cloud systems 31 and 32 and the in-house system 33. The plurality of virtual machines cooperate with each other to process information in response to a request from the user, and send a response to the user. For example, one virtual machine receives a request, a plurality of virtual machines transfer messages to a tree structure and perform data processing in a distributed manner, and the one virtual machine aggregates the final result and sends a response. The service administrator determines the placement of these multiple virtual machines.

The information collection server 100 is a server computer that monitors the usage status of resources in a plurality of virtual machines used for services. The information collection server 100 is located outside the cloud systems 31 and 32. The information collection server 100 may be owned by the service provider or may be included in the company's system 33.

The information collection server 100 periodically receives operation information indicating the resource usage status from at least a part of the plurality of virtual machines via the network 30. The resource usage is the usage of the hardware resources allocated to the virtual machine. Examples of the resource usage include CPU usage, RAM usage, storage access such as HDD, and network bandwidth usage. The information collection server 100 reports the resource usage status in the plurality of virtual machines to the service administrator. For example, the information collection server 100 displays a screen listing the resource usage status of a plurality of virtual machines.

The resource usage status of multiple virtual machines is referred to for service operation management. For example, configuration changes such as adding new virtual machines, deleting some virtual machines, increasing resources allocated to virtual machines, and changing the system in which virtual machines are placed are considered. The server 200 corresponds to the information processing device 20 of the first embodiment. The server 200b corresponds to the information processing device 20a of the first embodiment. The information collection server 100 corresponds to the information processing device 10 of the first embodiment.

FIG. 3 is a block diagram showing a hardware example of the information collection server.
The information collection server 100 includes a CPU 101, a RAM 102, an HDD 103, an image interface 104, an input interface 105, a medium reader 106, and a communication interface 107. These units included in the information collection server 100 are connected to the bus. The CPU 101 corresponds to the processing unit 13 of the first embodiment. The RAM 102 or the HDD 103 corresponds to the storage unit 12 of the first embodiment. The communication interface 107 corresponds to the communication unit 11 of the first embodiment. The servers 200, 200a, 200b, 200c, 200d, and 200e also have the same hardware as the information collection server 100.

The CPU 101 is a processor that executes a program instruction. The CPU 101 loads at least a part of the programs and data stored in the HDD 103 into the RAM 102 and executes the program. The CPU 101 may include a plurality of processor cores, and the information collection server 100 may include a plurality of processors. A collection of multiple processors is sometimes referred to as a "multiprocessor" or simply a "processor."

The RAM 102 is a volatile semiconductor memory that temporarily stores a program executed by the CPU 101 and data used by the CPU 101 for calculation. The information collection server 100 may include a type of memory other than RAM, or may include a plurality of memories.

The HDD 103 is a non-volatile storage that stores software programs such as an OS (Operating System), middleware, and application software, and data. The information collection server 100 may include other types of storage such as a flash memory and an SSD (Solid State Drive), or may include a plurality of storages.

The image interface 104 outputs an image to the display device 111 connected to the information collection server 100 in accordance with a command from the CPU 101. As the display device 111, any kind of display device such as a CRT (Cathode Ray Tube) display, a liquid crystal display (LCD), an organic EL (OEL: Organic Electro-Luminescence) display, and a projector can be used. .. An output device other than the display device 111 such as a printer may be connected to the information collection server 100.

The input interface 105 receives an input signal from the input device 112 connected to the information collection server 100. As the input device 112, any kind of input device such as a mouse, a touch panel, a touch pad, and a keyboard can be used. A plurality of types of input devices may be connected to the information collection server 100.

The medium reader 106 is a reading device that reads programs and data recorded on the recording medium 113. As the recording medium 113, any kind of recording medium such as a magnetic disk such as a flexible disk (FD) or HDD, an optical disk such as a CD (Compact Disc) or a DVD (Digital Versatile Disc), or a semiconductor memory is used. Can be done. The medium reader 106 copies, for example, a program or data read from the recording medium 113 to another recording medium such as the RAM 102 or the HDD 103. The read program is executed by, for example, the CPU 101. The recording medium 113 may be a portable recording medium, and may be used for distribution of programs and data. Further, the recording medium 113 and the HDD 103 may be referred to as a computer-readable recording medium.

The communication interface 107 is connected to the network 30 and communicates with the servers 200, 200a, 200b, 200c, 200d, 200e via the network 30. The communication interface 107 is a wired communication interface connected to, for example, a wired communication device such as a switch or a router. However, the communication interface 107 may be a wireless communication interface connected to a wireless communication device such as a base station or an access point.

Next, the arrangement and connection relationship of a plurality of virtual machines will be described.
FIG. 4 is a diagram showing an example of correspondence between a service, a virtual machine, and a system.
One example of a service operated in the second embodiment includes processes 41, 42, 43, 44, 45, 46. The topology of communication between processes 41, 42, 43, 44, 45, 46 is tree-structured. Process 41 receives a request from the user. Process 41 sends a message to processes 42 and 43, respectively, and requests processes 42 and 43 to process data. After receiving the message from the process 41, the process 43 sends a message to the process 44 to request the process 44 to process the data. After receiving the message from the process 43, the process 44 sends a message to the process 45 to request the process 45 to process data. After receiving the message from the process 41, the process 42 sends a message to the process 46 to request the process 46 to process data.

Process 45 performs the requested data processing. The process 44 receives the data processing result from the process 45 as a response to the transmitted message, and performs its own data processing. The process 43 receives the data processing result from the process 44 as a response to the transmitted message, and performs its own data processing. Process 46 performs the requested data processing. The process 42 receives the data processing result from the process 46 as a response to the transmitted message, and performs its own data processing. The process 41 receives the data processing result from the processes 42 and 43 as a response to the transmitted message. Finally, the process 41 processes its own data and sends a response to the user.

The process 41 is executed by the virtual machine 51. The process 42 is executed in the virtual machine 52. Process 43 is executed in virtual machine 53. Process 44 is executed in virtual machine 54. Process 45 is executed in virtual machine 55. Process 46 is executed in virtual machine 56. The processing procedure of processes 41, 42, 43, 44, 45, 46 is defined by the application program. The virtual machines 51, 52, 53, 54, 55, 56 of the second embodiment are virtual machines in a narrow sense that execute the guest OS and the application program. However, the virtual machines 51, 52, 53, 54, 55, 56 may be used as containers for container-type computer virtualization technology, and the virtual machines 51, 52, 53, 54, 55, 56 are not executing the guest OS. You may.

The virtual machines 51, 53, 54, 55 are arranged in the cloud system 31. In the cloud system 31, virtual machines 51, 53, 54, 55 may be arranged on different servers, or two or more of virtual machines 51, 53, 54, 55 are arranged on the same server. You may. The virtual machines 52 and 56 are arranged in the cloud system 32. In the cloud system 32, the virtual machines 52 and 56 may be arranged on different servers or may be arranged on the same server. In this example, the virtual machines 51, 52, 53, 54, 55, 56 are divided and arranged in two public cloud systems. However, it is also possible to arrange a part of the virtual machines 51, 52, 53, 54, 55, 56 in the in-house system 33. The virtual machine 51 corresponds to the virtual machine 21 of the first embodiment. The virtual machine 52 corresponds to the virtual machine 22 of the first embodiment.

Here, the information collection server 100 monitors the resource usage status of the virtual machines 51, 52, 53, 54, 55, and 56. In order to improve the monitoring accuracy of the resource usage status and ensure the real-time property of the monitoring information, it is preferable that the information collection server 100 receives the operation information with high frequency. However, if the information collection server 100 receives operation information from all of the virtual machines 51, 52, 53, 54, 55, and 56 at high frequency, there is a problem that the traffic of the network 30, which is a wide area data communication network, increases.

Therefore, the information processing system of the second embodiment stops the transmission of operation information from some of the virtual machines 51, 52, 53, 54, 55, 56 to the information collection server 100. The virtual machine that stops sending operation information is a virtual machine whose resource usage has a strong correlation with other specific virtual machines. The resource usage status of the virtual machine that does not collect the operation information is estimated based on the correlation calculated in advance with the resource usage status of the virtual machine that the information collection server 100 collects the operation information. As a result, the traffic of the network 30 can be reduced while maintaining the monitoring accuracy.

FIG. 5 is a sequence diagram showing an example of communication between the information collection server and the virtual machine.
Here, among the virtual machines 51, 52, 53, 54, 55, 56, the communication of the virtual machines 51, 53 will be described as a representative. In the sequence example of FIG. 5, the virtual machine 51 continues to transmit the operation information to the information collection server 100, and the virtual machine 53 temporarily stops the transmission of the operation information to the information collection server 100.

First, the information collection server 100 transmits a test start notification to the virtual machines 51 and 53 (S100). Upon receiving the test start notification, the virtual machines 51 and 53 start monitoring communication with other virtual machines. The communication monitoring may be packet-level transmission packet monitoring or application-level transmission message monitoring. The virtual machines 51 and 53 collect a communication log including at least the destination and time of the transmitted message.

The information collection server 100 transmits a test request to the virtual machine 51, which is the entrance of the service (S101). Upon receiving the test request, the virtual machine 51 sends a test message to the virtual machine 53 to request the virtual machine 53 to process data (S102). At this time, the virtual machine 51 detects the transmission of the test message to the virtual machine 53, and collects the communication log indicating the transmission of the test message to the virtual machine 53. When the processing of the test request is completed, the virtual machines 51 and 53 transmit the communication log to the information collection server 100 (S103). The virtual machines 51 and 53 autonomously transmit the communication log. However, the information collection server 100 may request the communication logs from the virtual machines 51 and 53.

The information collection server 100 also receives communication logs from other virtual machines. The information collection server 100 determines the order of message transmission between virtual machines based on these communication logs, and determines the data flow of the service as shown in FIG. Further, the information collection server 100 determines a low collection frequency as an initial value of the operation information collection frequency. The low collection frequency is, for example, 1 minute intervals. The information collection server 100 may determine the initial value of the collection frequency according to the total number of virtual machines participating in the service. In that case, the smaller the total number of virtual machines, the higher the collection frequency, and the larger the total number of virtual machines, the lower the collection frequency. Before the group classification described later, the information collection server 100 collects the operation information at a low collection frequency, so that it is possible to prevent the communication volume of the entire information processing system from becoming excessive temporarily.

The information collection server 100 notifies the virtual machines 51 and 53 of the initial value of the collection frequency (S104). The virtual machines 51 and 53 start acquiring the resource usage status and transmitting the operation information to the information collection server 100 at a low frequency notified from the information collection server 100 (S105). Examples of the resource usage include CPU usage, RAM usage, HDD access, network bandwidth usage, and the like. In the second embodiment, the CPU and RAM are mainly focused on as resources.

The information collection server 100 retains the operation information received from the virtual machines 51 and 53 as a history for at least a certain period of time. When the history of operation information for a certain period of time is accumulated, the information collection server 100 classifies the plurality of virtual machines into groups based on the history of operation information and the data flow between the plurality of virtual machines. When two virtual machines have a parent-child relationship on the data flow and there is a strong correlation between the resource usage status of the two virtual machines, the two virtual machines are classified into the same group. When the absolute value of the correlation coefficient of the measured value included in the operation information is larger than a predetermined threshold value (for example, 0.8) between the two virtual machines, it is determined that there is a strong correlation. A parent-child relationship is one in which one virtual machine sends a message to the other virtual machine. According to the example of FIG. 4, focusing on the virtual machines 51 and 53, the virtual machine 51 is a parent virtual machine and the virtual machine 53 is a child virtual machine.

If the resource usage of the grandchild virtual machine that has a parent-child relationship with the child virtual machine has a strong correlation with the parent virtual machine, the grandchild virtual machine is also continuously classified into the same group as the parent virtual machine. The information collection server 100 selects the most upstream virtual machine in each group as the representative virtual machine, and treats the virtual machines other than the representative virtual machine as the subordinate virtual machines. As will be described later, only the representative virtual machine transmits the operation information to the information collection server 100, and the subordinate virtual machine does not transmit the operation information to the information collection server 100. The information collection server 100 may limit the virtual machines classified into the same group to the virtual machines arranged in the same system among the cloud systems 31 and 32 and the in-house system 33.

Virtual machines are grouped by resource type. Therefore, two virtual machines may be classified into the same group because they have a strong correlation with respect to CPU usage, while they may be classified into different groups because they have a low correlation with respect to RAM usage. Here, it is assumed that the virtual machine 51 is the representative virtual machine and the virtual machine 53 is the subordinate virtual machine. In addition, the information collection server 100 calculates a magnification for the resource usage status of the representative virtual machine for the subordinate virtual machine. The magnification can be calculated, for example, by dividing the average of the measured values of the dependent virtual machines for a certain period by the average of the measured values of the representative virtual machine. However, the magnification may be expressed as a function using a variable representing the measured value of the representative virtual machine.

The information collection server 100 transmits a group notification to the virtual machine 51, which is the representative virtual machine (S106). The group notification includes the identification information of the virtual machine 53, which is a subordinate virtual machine, and the magnification of the measured value of the virtual machine 53 with respect to the measured value of the virtual machine 51. Upon receiving the group notification, the virtual machine 51 transmits a stop notification to the virtual machine 53, which is a subordinate virtual machine belonging to the same group as the virtual machine 51 (S107). Upon receiving the stop notification, the virtual machine 53 stops transmitting the operation information to the information collection server 100.

Further, the virtual machine 51 changes the frequency of transmitting the operation information to the information collection server 100 to be higher than the initial value. The virtual machine 51 starts to acquire the resource usage status and transmit the operation information to the information collection server 100 at a high collection frequency after the change (S108).

The changed collection frequency depends on the number of dependent virtual machines that belong to the same group. Assuming that the total number of virtual machines in the group is n and the number of dependent virtual machines is m, the transmission frequency of the representative virtual machine is increased so as to be n ÷ (nm) × α times the initial value. That is, the transmission interval of the representative virtual machine is shortened so as to be (nm) ÷ n ÷ α times the initial value. The more dependent virtual machines that do not send operation information, the higher the frequency of sending the representative virtual machine. The adjustment parameter α is a positive real number less than 1. As a result, it is possible to increase the collection frequency of representative virtual machines while reducing the total amount of operation information in the group.

However, if the collection frequency is too high, it does not contribute to the improvement of monitoring accuracy and the network load is unnecessarily increased. Therefore, the collection frequency of the representative virtual machine should not be higher than the predetermined upper limit. The upper limit of the collection frequency is, for example, an interval of 1 second. When two virtual machines of virtual machines 51 and 53 form one group, for example, the collection frequency of the virtual machine 51, which is a representative virtual machine, is adjusted between 30 seconds and 1 minute.

The information collection server 100 receives the operation information from the virtual machine 51, but does not receive the operation information from the virtual machine 53. Therefore, the information collection server 100 multiplies the measured value indicated by the latest operation information received from the virtual machine 51 by the magnification of the virtual machine 53 calculated in advance, so that the latest resource usage status of the virtual machine 53 is used. To estimate. The calculation of the magnification and the estimation of the resource usage status are performed for each resource type.

In addition, the virtual machine 51 monitors the resource usage status of the virtual machine 53, which is a subordinate virtual machine belonging to the same group. The virtual machine 51 requests operation information from the virtual machine 53 at a low predetermined frequency (S109). The low predetermined frequency may be the same as the initial value of the collection frequency notified in step S104, for example, at 1-minute intervals. The virtual machine 53 acquires the resource usage status in the virtual machine 53 in response to the request from the virtual machine 51, and transmits the operation information to the virtual machine 51 (S110).

When the virtual machine 51 receives the operation information from the virtual machine 53, the virtual machine 51 determines whether a strong correlation between the virtual machine 51 and the virtual machine 53 in the resource usage status is maintained. Whether or not the strong correlation is maintained is determined by comparing the latest operation information of the virtual machine 53 with the operation information of the corresponding virtual machine 51.

For example, the virtual machine 51 extracts the operation information of the virtual machine 51 closest to the reception time of the operation information from the virtual machine 53 from the history, and the measured value included in the operation information of the virtual machine 51 and the operation information of the virtual machine 53. Create a pair with the measured value included in. The virtual machine 51 calculates the correlation coefficient from a plurality of pairs, determines that a strong correlation is maintained when the absolute value of the correlation coefficient is larger than the threshold value, and the absolute value of the correlation coefficient is equal to or less than the threshold value. In that case, it is judged that the correlation is low. Further, when the virtual machine 51 determines that the strong correlation is maintained, the virtual machine 51 recalculates the ratio of the resource usage status of the virtual machine 53 to the virtual machine 51. For example, the virtual machine 51 is calculated by dividing the average of the measured values of the most recent plurality of virtual machines 53 by the average of the measured values of the corresponding virtual machine 51.

Here, it is assumed that a strong correlation is maintained between the virtual machine 51 and the virtual machine 53. Then, the virtual machine 51 transmits a correlation update notification including the recalculated magnification to the information collection server 100 (S111). Further, the virtual machine 51 continues to acquire the resource usage status and transmits the operation information to the information collection server 100 at a high collection frequency (S112). The information collection server 100 does not continue to receive operation information from the virtual machine 53. Therefore, the information collection server 100 estimates the resource usage status in the virtual machine 53 by using the latest magnification notified in step S111.

Again, the virtual machine 51 requests the virtual machine 53 for operation information (S113), and receives the operation information from the virtual machine 53 (S114). Here, it is assumed that the absolute value of the correlation coefficient becomes equal to or less than the threshold value and the correlation between the virtual machine 51 and the virtual machine 53 is lost. Then, the virtual machine 51 transmits a restart notification to the virtual machine 53 (S115). Upon receiving the resume notification, the virtual machine 53 periodically acquires the resource usage status and resumes transmitting the resource usage status to the information collection server 100. The collection frequency of the virtual machine 53 at this time may be low, and is, for example, the initial value notified in step S104.

The virtual machine 51 transmits a group release notification indicating that the virtual machine 53 has left the group to the information collection server 100 (S116). When the information collection server 100 receives the group release notification, the information collection server 100 switches to grasp the resource usage status of the virtual machine 53 out of the group based on the measured value instead of the estimated value.

Further, the virtual machine 51 recalculates the frequency with which the virtual machine 51 transmits the operation information to the information collection server 100 as the number of dependent virtual machines belonging to the group decreases. As a result, the collection frequency of the virtual machine 51 is reduced. As a result, the virtual machines 51 and 53 frequently acquire the resource usage status and transmit the operation information to the information collection server 100 (S117). The collection frequency returns, for example, to 1 minute intervals.

In this way, there is a strong correlation between the parent virtual machine that sends a message in the service and the child virtual machine that receives the message and processes data, in terms of resource usage such as CPU and RAM. There is. In this case, the traffic of the network 30 can be reduced by stopping the collection of the operation information of the child virtual machine and estimating the resource usage status of the child virtual machine from the operation information of the parent virtual machine. Further, by increasing the collection frequency of the parent virtual machine within the range in which the total traffic of the network 30 is reduced, the resource usage status can be monitored in real time, and the monitoring accuracy can be improved.

The information collection server 100 periodically determines the correlation between the representative virtual machines that are collecting the operation information even after the collection of the operation information of the subordinate virtual machines is stopped. When a strong correlation is newly detected between the two representative virtual machines, the information collection server 100 changes the representative virtual machine on the downstream side of the data flow to the dependent virtual machine. In that case, the information collection server 100 transmits a group notification indicating that the subordinate virtual machine has been added to the representative virtual machine on the upstream side, as in step S106.

FIG. 6 is a diagram showing an example of grouping virtual machines.
As described above, the information collection server 100 transmits a test request to the virtual machine 51. Then, according to the structure shown in FIG. 4, the virtual machine 51 sends a test message to the virtual machines 52 and 53. The virtual machine 53 sends a test message to the virtual machine 54, and the virtual machine 54 sends a test message to the virtual machine 55. The virtual machine 52 sends a test message to the virtual machine 56. During this time, the virtual machines 51, 52, 53, 54, 55, 56 collect a communication log including the virtual machine to which the test message is sent and the transmission time of the test message by a method such as monitoring packets.

The virtual machines 51, 52, 53, 54, 55, 56 transmit the communication log to the information collection server 100. The information collection server 100 determines the order of message transmission between the virtual machines 51, 52, 53, 54, 55, and 56 based on the received communication log, and grasps the data flow. Further, the information collection server 100 uses the virtual machines 51, 52, 53, 54, 55, and 56 at low frequency (for example, at 1-minute intervals) to obtain CPU information indicating the CPU usage rate and RAM usage amount. Receives both RAM information, which is the operation information shown.

FIG. 7 is a diagram (continuation 1) showing an example of grouping virtual machines.
The information collection server 100 groups virtual machines 51, 52, 53, 54, 55, 56 for each resource type, that is, for each of the CPU and RAM.

Regarding the CPU, the information collection server 100 uses the CPU information received from the virtual machines 51, 52, 53, 54, 55, and 56 infrequently to correlate from the virtual machine 51 at the head of the data flow toward the downstream. Judge in order.

First, the information collection server 100 calculates the correlation coefficient of the CPU usage rate between the virtual machine 51 and the virtual machine 53. Here, it is assumed that the absolute value of the correlation coefficient is larger than the threshold value. Then, the information collection server 100 classifies the virtual machine 53 into the same group as the virtual machine 51. Next, the information collection server 100 traces the data flow downstream and calculates the correlation coefficient of the CPU usage rate between the virtual machine 51 and the virtual machine 54. Here, it is assumed that the absolute value of the correlation coefficient is larger than the threshold value. Then, the information collection server 100 classifies the virtual machine 54 into the same group as the virtual machines 51 and 53.

Next, the information collection server 100 traces the data flow downstream and calculates the correlation coefficient of the CPU usage rate between the virtual machine 51 and the virtual machine 55. Here, it is assumed that the absolute value of the correlation coefficient is equal to or less than the threshold value. Then, the information collection server 100 classifies the virtual machine 55 into a different group from the virtual machines 51, 53, 54. In addition, the information collection server 100 calculates the correlation coefficient of the CPU usage rate between the virtual machine 51 and the virtual machine 52. Here, it is assumed that the absolute value of the correlation coefficient is equal to or less than the threshold value. Then, the information collection server 100 classifies the virtual machine 52 into a different group from the virtual machine 51. Next, the information collection server 100 traces the data flow downstream and calculates the correlation coefficient of the CPU usage rate between the virtual machine 52 and the virtual machine 56. Here, it is assumed that the absolute value of the correlation coefficient is larger than the threshold value. Then, the information collection server 100 classifies the virtual machine 56 into the same group as the virtual machine 52.

As a result, with respect to the CPU, the virtual machines 51, 52, 55 become representative virtual machines, and the virtual machines 53, 54, 56 become subordinate virtual machines. The virtual machines 51, 52, and 55 periodically transmit CPU information to the information collection server 100. The virtual machines 53, 54, and 56 stop transmitting CPU information to the information collection server 100. Further, as the virtual machines 53 and 54 stop transmitting CPU information, the collection frequency of the virtual machines 51 increases with the upper limit being three times the initial value (for example, the lower limit of the transmission interval being 20 seconds). Further, as the virtual machine 56 stops transmitting CPU information, the collection frequency of the virtual machine 52 increases with the upper limit being twice the initial value (for example, the lower limit of the transmission interval being 30 seconds).

Similarly, with respect to the RAM, the information collection server 100 uses the RAM information received from the virtual machines 51, 52, 53, 54, 55, 56 infrequently from the virtual machine 51 at the head of the data flow toward the downstream. Determine the correlation in order.

First, the information collection server 100 calculates the correlation coefficient of the RAM usage between the virtual machine 51 and the virtual machine 53. Here, it is assumed that the absolute value of the correlation coefficient is larger than the threshold value. Then, the information collection server 100 classifies the virtual machine 53 into the same group as the virtual machine 51. Next, the information collection server 100 traces the data flow downstream and calculates the correlation coefficient of the RAM usage between the virtual machine 51 and the virtual machine 54. Here, it is assumed that the absolute value of the correlation coefficient is equal to or less than the threshold value. Then, the information collection server 100 classifies the virtual machine 54 into a different group from the virtual machines 51 and 53.

Next, the information collection server 100 traces the data flow downstream and calculates the correlation coefficient of the RAM usage between the virtual machine 54 and the virtual machine 55. Here, it is assumed that the absolute value of the correlation coefficient is larger than the threshold value. Then, the information collection server 100 classifies the virtual machine 55 into the same group as the virtual machine 54. In addition, the information collection server 100 calculates the correlation coefficient of the RAM usage between the virtual machine 51 and the virtual machine 52. Here, it is assumed that the absolute value of the correlation coefficient is equal to or less than the threshold value. Then, the information collection server 100 classifies the virtual machine 52 into a different group from the virtual machine 51. Next, the information collection server 100 traces the data flow downstream and calculates the correlation coefficient of the RAM usage between the virtual machine 52 and the virtual machine 56. Here, it is assumed that the absolute value of the correlation coefficient is equal to or less than the threshold value. Then, the information collection server 100 classifies the virtual machine 56 into a group different from the virtual machine 52.

As a result, with respect to RAM, virtual machines 51, 52, 54, and 56 become representative virtual machines, and virtual machines 53 and 55 become subordinate virtual machines. The virtual machines 51, 52, 54, and 56 periodically transmit RAM information to the information collection server 100. The virtual machines 53 and 55 stop transmitting RAM information to the information collection server 100. Further, as the virtual machine 53 stops transmitting RAM information, the collection frequency of the virtual machine 51 increases up to twice the initial value. Further, as the virtual machine 55 stops transmitting RAM information, the collection frequency of the virtual machine 54 increases up to twice the initial value.

FIG. 8 is a diagram (continuation 2) showing an example of grouping virtual machines.
Here, the group change of CPU information will be described. The RAM information is group-changed independently of the CPU information. The virtual machine 51, which is a representative virtual machine, frequently requests CPU information from virtual machines 53, 54, which are subordinate virtual machines belonging to the same group, and receives CPU information from the virtual machines 53, 54. Further, the virtual machine 52, which is a representative virtual machine, frequently requests CPU information from the virtual machine 56, which is a subordinate virtual machine belonging to the same group, and receives the CPU information from the virtual machine 56.

The virtual machine 51 calculates the correlation coefficient of the CPU usage rate between the virtual machine 51 and the virtual machine 53, and calculates the correlation coefficient of the CPU usage rate between the virtual machine 51 and the virtual machine 54. Here, it is assumed that the absolute value of any correlation coefficient is larger than the threshold value. Then, the virtual machines 53 and 54 continue to operate as subordinate virtual machines with respect to the virtual machine 51. In addition, the virtual machine 52 calculates the correlation coefficient of the CPU usage rate between the virtual machine 52 and the virtual machine 56. Here, it is assumed that the absolute value of the correlation coefficient drops below the threshold value. Then, the virtual machine 56 is removed from the group of the virtual machine 52 and operates as a representative virtual machine. Therefore, the virtual machine 56 restarts the transmission of the CPU information to the information collection server 100. Further, as the transmission of the virtual machine 56 is resumed, the collection frequency of the virtual machine 52 decreases.

On the other hand, the information collection server 100 monitors the correlation of CPU information periodically received from the virtual machines 51, 52, 55. The information collection server 100 selects the virtual machine 51 and the virtual machines 55 adjacent to the group according to the data flow. The information collection server 100 calculates the correlation coefficient of the CPU usage rate between the virtual machine 51 and the virtual machine 55. Here, it is assumed that the absolute value of the correlation coefficient becomes larger than the threshold value. Then, the virtual machine 55 enters the group of the virtual machines 51, 53, 54 and operates as a subordinate virtual machine. Therefore, the virtual machine 55 stops transmitting CPU information to the information collection server 100.

Further, the information collection server 100 selects the virtual machine 51 and the virtual machines 52 adjacent to the group according to the data flow. The information collection server 100 calculates the correlation coefficient of the CPU usage rate between the virtual machine 51 and the virtual machine 52. Here, it is assumed that the absolute value of the correlation coefficient is still below the threshold value. Then, the virtual machine 52 continues to operate as a representative virtual machine, and continues to transmit CPU information to the information collection server 100.

In this way, the existing representative virtual machine may be changed to the subordinate virtual machine at the discretion of the information collection server 100. In addition, the existing subordinate virtual machine may be changed to the representative virtual machine at the discretion of the representative virtual machine.

Next, the functions of the information collection server 100 and each virtual machine will be described.
FIG. 9 is a block diagram showing a functional example of the information collection server and the virtual machine.
The information collection server 100 has a structural information storage unit 121, an operation history storage unit 122, and a correlation information storage unit 123. These storage units are realized by using, for example, the storage area of the RAM 102 or the HDD 103. Further, the information collection server 100 includes a test execution unit 124, an operation information reception unit 125, a correlation detection unit 126, a correlation update unit 127, and an operation status display unit 128. These processing units are realized, for example, by using a program.

The structure information storage unit 121 stores service structure information representing a data flow between a plurality of virtual machines used for the service. The service structure information represents a topology in which multiple virtual machines are connected to a tree structure. The operation history storage unit 122 stores the history of operation information received from the virtual machines 51, 52, 53, 54, 55, 56. The operation history storage unit 122 records the measured value in association with the time, the virtual machine, and the resource set. The operation history storage unit 122 holds the received operation information as a history for at least a certain period of time. The correlation information storage unit 123 stores the correlation information indicating the relationship between the representative virtual machine and the dependent virtual machine having a strong correlation in the resource usage status. The correlation information includes a magnification for estimating the numerical value of the dependent virtual machine from the measured value of the representative virtual machine.

The test execution unit 124 executes a test for determining the data flow of the virtual machines 51, 52, 53, 54, 55, 56. The test execution unit 124 notifies the virtual machines 51, 52, 53, 54, 55, and 56 of the start of the test, and sends a test request to the virtual machine 51, which is the entrance of the service. The test execution unit 124 collects communication logs indicating the virtual machine to which the test message is sent and the transmission time from the virtual machines 51, 52, 53, 54, 55, 56, analyzes the communication log, and determines the data flow. To do. The test execution unit 124 generates structural information indicating a data flow and stores it in the structural information storage unit 121.

The operation information receiving unit 125 receives operation information periodically transmitted by at least a part of the virtual machines 51, 52, 53, 54, 55, 56. The operation information is divided according to the type of resource, and which resource the operation information is transmitted depends on the virtual machine. In the initial state, all of the virtual machines 51, 52, 53, 54, 55, 56 transmit operation information about all resources at low frequency. The operation information receiving unit 125 overwrites the correlation information storage unit 123 with the latest measured value included in the received operation information. Further, the operation information receiving unit 125 associates the measured value included in the received operation information with the identification information of the source virtual machine, the reception time, and the set of resources indicated by the measured value, and stores the measured value in the operation history storage unit 122. Append.

The correlation detection unit 126 periodically transmits operation information for each resource based on the structural information stored in the structural information storage unit 121 and the history of the operation information stored in the operation history storage unit 122. Determine the correlation between virtual machines. The correlation detection unit 126 designates the upstream virtual machine as the representative virtual machine and the downstream virtual machine as the dependent virtual machine among the two virtual machines in which the strong correlation is detected. In addition, the correlation detection unit 126 calculates the ratio of the resource usage status of the subordinate virtual machine to the representative virtual machine. The correlation detection unit 126 records the relationship and magnification between the representative virtual machine and the dependent virtual machine in the correlation information storage unit 123.

Specifically, the correlation detection unit 126 selects the resource type, selects two virtual machines that are upstream and downstream of the data flow, and the resource and the two virtual machines for the most recent fixed period of time. The operation information is extracted from the operation history storage unit 122. The correlation detection unit 126 associates the measured value of one virtual machine with the measured value of the other virtual machine having the closest reception time, and calculates the correlation coefficient from a plurality of measured value pairs at different timings. The correlation detection unit 126 determines that the correlation is strong when the absolute value of the correlation coefficient is larger than the threshold value, and determines that the correlation is weak when the absolute value of the correlation coefficient is equal to or less than the threshold value. When the correlation is strong, the correlation detection unit 126 calculates the average of the measured values of the representative virtual machine for the most recent fixed period and the average of the measured values of the dependent virtual machine for the most recent fixed period, and divides the latter by the former. Calculate the magnification.

The correlation update unit 127 receives a correlation update notification indicating that the magnification has been updated from the representative virtual machine. Then, the correlation update unit 127 records the updated magnification in the correlation information storage unit 123. In addition, the correlation update unit 127 receives a group release notification from the representative virtual machine indicating that a certain subordinate virtual machine has left the group. Then, the correlation update unit 127 changes the correlation recorded in the correlation information storage unit 123.

The operation status display unit 128 displays a management screen listing the resource usage status of the virtual machines 51, 52, 53, 54, 55, and 56 on the display device 111 in response to a request from the user. At this time, the operation status display unit 128 may display the measured value for the set of the virtual machine and the resource in which the measured value is recorded in the correlation information storage unit 123. On the other hand, the operation status display unit 128 calculates an estimated value by multiplying the measured value of the related representative virtual machine by a magnification for the set of the virtual machine and the resource for which the measured value is not recorded in the correlation information storage unit 123. Display the estimated value instead of the measured value. The operation status display unit 128 may continuously display the management screen without waiting for a request from the user. Further, the operation status display unit 128 may store the management information enumerating the resource usage status in a storage device such as the HDD 103, or may transmit the management information to another information processing device.

The virtual machine 51 has a group information storage unit 221. The group information storage unit 221 is realized by using, for example, a storage area of a RAM or an HDD. Further, the virtual machine 51 has a service processing unit 222, a communication log acquisition unit 223, an operation information acquisition unit 224, and a group control unit 225. These processing units are realized, for example, by using a program. Other virtual machines also have the same functions as the virtual machine 51.

The group information storage unit 221 stores group information for managing subordinate virtual machines belonging to the same group when the virtual machine 51 is a representative virtual machine. The group information includes the identification information of the subordinate virtual machine notified from the information collection server 100. In addition, the group information includes the measured values collected from the subordinate virtual machines. In addition, the group information includes the magnification notified from the information collection server 100 or the latest updated magnification.

The service processing unit 222 executes the data processing of the service according to the application program. When the virtual machine 51 corresponds to the entrance of the service, the service processing unit 222 receives the test request from the information collection server 100. The service processing unit 222 may send a message to another virtual machine. In addition, the service processing unit 222 may receive a message from another virtual machine.

When the communication log acquisition unit 223 receives the test start notification from the information collection server 100, the communication log acquisition unit 223 monitors the packet transmitted by the virtual machine 51. Alternatively, the communication log acquisition unit 223 monitors the application-level transmitted message. The communication log acquisition unit 223 detects the transmission destination virtual machine and the transmission time of the message at the time of test execution, generates a communication log indicating the transmission destination virtual machine and the transmission time, and transmits the communication log to the information collection server 100.

The operation information acquisition unit 224 acquires a measured value indicating the current resource usage status in the virtual machine 51. Examples of the measured value include CPU usage rate, RAM usage amount, HDD access amount, network bandwidth usage rate, and the like. When the guest OS is executed in the virtual machine 51, the operation information acquisition unit 224 may acquire the measured value from the guest OS.

When the information collection server 100 notifies the initial value of the collection frequency, the operation information acquisition unit 224 acquires the measured values of all resources at the specified collection frequency and transmits them to the information collection server 100. When the group control unit 225 determines that the collection frequency of a certain resource is changed, the operation information acquisition unit 224 acquires and transmits the measured value at the changed collection frequency. When a stop notification is received from another virtual machine for a certain resource, the operation information acquisition unit 224 stops the acquisition and transmission of the measured value. When a restart notification is received from another virtual machine for a certain resource, the operation information acquisition unit 224 resumes the acquisition and transmission of the measured value. When operation information is requested from another virtual machine for a certain resource, the operation information acquisition unit 224 acquires the measured value and returns it to the other virtual machine.

The group control unit 225 manages the subordinate virtual machine when the virtual machine 51 is the representative virtual machine. The group control unit 225 records the identification information and the magnification of the subordinate virtual machine notified from the information collection server 100 in the group information storage unit 221. In addition, the group control unit 225 sends a stop notification to the subordinate virtual machine. Further, the group control unit 225 changes the collection frequency of the virtual machines 51 according to the number of dependent virtual machines.

Further, the group control unit 225 periodically requests the operation information of the subordinate virtual machine and collects the operation information of the subordinate virtual machine. The group control unit 225 determines whether a strong correlation is maintained between the operation information of the virtual machine 51 and the operation information of the subordinate virtual machine. When a strong correlation is maintained, the group control unit 225 calculates the latest magnification, updates the magnification recorded in the group information storage unit 221 and sends a correlation update notification including the latest magnification to the information collection server 100. Send to. If the strong correlation is not maintained, the group control unit 225 sends a restart notification to the subordinate virtual machine, updates the group information of the group information storage unit 221 and sends a group release notification to the information collection server 100.

FIG. 10 is a diagram showing an example of a service structure table and an operation history table.
The service structure table 131 is stored in the structure information storage unit 121. The service structure table 131 includes items for the virtual machine ID, the parent virtual machine, and the child virtual machine. The virtual machine identification information is registered in the virtual machine ID item. In the item of the parent virtual machine, the identification information of the upstream virtual machine that is the source of sending the message to the virtual machine is registered. In the child virtual machine item, the identification information of the downstream virtual machine to which the virtual machine sends a message is registered.

The service structure table 131 represents a tree-structured data flow between virtual machines 51, 52, 53, 54, 55, 56 as shown in FIG. For example, the child virtual machines of the virtual machine 51 (VM1) are the virtual machine 52 (VM2) and the virtual machine 53 (VM3). The parent virtual machine of the virtual machine 52 (VM2) is the virtual machine 51 (VM1). The child virtual machine of the virtual machine 52 (VM2) is the virtual machine 56 (VM6).

The operation history table 132 is stored in the operation history storage unit 122. The operation history table 132 includes items of time, virtual machine ID, resource, and measured value. In the time item, the time when the information collection server 100 receives the operation information is registered. However, as the time, the time included in the operation information may be used, or the time when the virtual machine acquires the measured value or the time when the operation information is transmitted may be used. In the virtual machine ID item, the identification information of the virtual machine that sent the operation information is registered. In the resource item, the resource type indicated by the measured value is registered. Examples of resource types include CPU, RAM, storage, and network. The measured value included in the operation information is registered in the measured value item.

FIG. 11 is a diagram showing an example of a correlation table and an operation status table.
The correlation table 133 is stored in the correlation information storage unit 123. The correlation table 133 includes items such as virtual machine ID, resource, measured value, dependent virtual machine, and magnification. In the virtual machine ID item, the identification information of the representative virtual machine that is transmitting the operation information is registered. In the resource item, the type of resource such as CPU and RAM is registered. In the measurement value item, the measurement value included in the latest operation information received for the representative virtual machine and the resource is registered. In the item of the dependent virtual machine, the identification information of the dependent virtual machine having a strong correlation with the representative virtual machine for the resource is registered. As the magnification, the magnification of the numerical value of the dependent virtual machine with respect to the measured value of the representative virtual machine is registered.

For example, in the correlation table 133, the latest CPU measurement value of the virtual machine 51 is 50, the CPU magnification of the virtual machine 53 with respect to the virtual machine 51 is 0.50, and the virtual machine 54 with respect to the virtual machine 51. It is registered that the magnification of the CPU is 0.30. Further, for example, it is registered that the latest measured value of the RAM of the virtual machine 51 is 30, and the magnification of the RAM of the virtual machine 53 with respect to the virtual machine 51 is 0.40. Further, for example, it is registered that the latest measured value of the CPU of the virtual machine 52 is 70, and the magnification of the CPU of the virtual machine 56 with respect to the virtual machine 52 is 1.20.

The operation status table 134 is generated by the operation status display unit 128 and displayed on the display device 111. The operation status table 134 includes items of virtual machine ID, resource, and current value. The virtual machine identification information is registered in the virtual machine ID item. In the resource item, the type of resource such as CPU and RAM is registered. In the current value item, a numerical value indicating the latest resource usage status is registered. The numerical value registered in the current value item is the latest measured value for the representative virtual machine and the estimated value for the dependent virtual machine. It may be added that the operation status table 134 is an estimated value.

For example, the measured value "50" is displayed as the CPU usage rate of the virtual machine 51. Further, the measured value "30" is displayed as the RAM usage amount of the virtual machine 51. On the other hand, as the CPU usage rate of the virtual machine 53, "25", which is an estimated value calculated by multiplying 50, which is the measured value of the CPU of the virtual machine 51, by 0.50, is displayed. Further, as the amount of RAM used by the virtual machine 53, "12", which is an estimated value calculated by multiplying 30 which is the measured value of the RAM of the virtual machine 51 by 0.40, is displayed.

FIG. 12 is a diagram showing an example of a group management table.
The group management table 231 is stored in the group information storage unit 221. The group management table 231 includes items of resources, self-measured values, dependent virtual machines, other measured values, and magnifications. In the resource item, the type of resource such as CPU and RAM is registered. In the self-measured value item, a numerical value indicating the usage status of the resource measured by the virtual machine 51 is registered. In the item of the subordinate virtual machine, the identification information of the subordinate virtual machine belonging to the same group as the virtual machine 51 is registered. In the item of other measured values, a numerical value indicating the usage status of the resource measured by the dependent virtual machine is registered. In the item of magnification, the latest average magnification of the measured value of the dependent virtual machine with respect to the measured value of the virtual machine 51 is registered.

Next, the processing procedure of the information collection server 100 and the representative virtual machine will be described.
FIG. 13 is a flowchart showing a procedure example of the server start process.
(S10) The test execution unit 124 transmits a test request to a predetermined virtual machine. A message is sent between multiple virtual machines in response to a test request.

(S11) The test execution unit 124 receives the communication log from each virtual machine. The communication log contains the virtual machine to which the message is sent and the transmission time.
(S12) The test execution unit 124 determines the data flow between the plurality of virtual machines based on the communication log received in step S11, and generates the service structure table 131.

(S13) The operation information receiving unit 125 determines the initial value of the collection frequency for collecting the operation information according to the total number of virtual machines participating in the service.
(S14) The operation information receiving unit 125 notifies each virtual machine of the collection frequency. As a result, the operation information receiving unit 125 receives the operation information from each virtual machine at the notified frequency. The operation information receiving unit 125 stores the received operation information in the operation history table 132.

(S15) The correlation detection unit 126 groups a plurality of virtual machines for each resource based on the service structure table 131 and the operation history table 132. One virtual machine in the group becomes a representative virtual machine that transmits operation information, and another virtual machine becomes a subordinate virtual machine that does not transmit operation information. Details of group determination will be described later.

(S16) The correlation detection unit 126 notifies the representative virtual machine determined in step S15 of the identification information of the subordinate virtual machines belonging to the same group and the magnification thereof. Since grouping is performed for each resource, the notification is also performed for each resource.

FIG. 14 is a flowchart showing an example of a procedure for group determination.
The group determination is executed in step S15 described above and step S64 described later.
(S20) The correlation detection unit 126 selects one resource. The resources to be monitored are predetermined such as CPU, RAM, HDD, and network.

(S21) The correlation detection unit 126 selects a root (starting point) virtual machine from the tree-structured data flow shown in the service structure table 131, and designates it as a representative virtual machine.
(S22) The correlation detection unit 126 determines whether or not a virtual machine for which operation information is being collected remains in the tree-structured data flow. If the corresponding virtual machine remains, the process proceeds to step S23, and if the corresponding virtual machine does not remain, the process proceeds to step S29.

(S23) The correlation detection unit 126 traces the data flow downstream from the currently selected virtual machine and selects the next virtual machine. When the data flow is branched, one of the branching directions may be followed. When the end of the tree structure is reached, the correlation detection unit 126 returns to the immediately preceding branch point and follows the unselected branch direction.

(S24) The correlation detection unit 126 extracts the operation history of the resource selected in step S20 for the virtual machine selected in step S23 from the operation history table 132. Further, the correlation detection unit 126 extracts the operation history of the resource selected in step S20 for the immediately preceding representative virtual machine on the upstream side of the virtual machine from the operation history table 132. The correlation detection unit 126 calculates the correlation coefficient of the measured value included in the extracted operation information between the immediately preceding representative virtual machine and the selected virtual machine.

(S25) The correlation detection unit 126 compares the absolute value of the correlation coefficient calculated in step S24 with a predetermined threshold value. If the absolute value of the correlation coefficient is greater than the threshold value, the process proceeds to step S26, and if the absolute value of the correlation coefficient is equal to or less than the threshold value, the process proceeds to step S28.

(S26) The correlation detection unit 126 classifies the virtual machine selected in step S23 into the same group as the immediately preceding representative virtual machine and designates it as a dependent virtual machine.
(S27) The correlation detection unit 126 calculates the ratio of the measured value of the dependent virtual machine to the measured value of the representative virtual machine by using the operation information extracted from the operation history table 132 in step S24. The correlation detection unit 126 registers in the correlation table 133 that the representative virtual machine and the dependent virtual machine have a strong correlation and the magnification. Then, the process returns to step S22.

(S28) The correlation detection unit 126 designates the virtual machine selected in step S23 as the representative virtual machine. Then, the process returns to step S22.
(S29) The correlation detection unit 126 determines whether all the resources have been selected in step S20. When all resources are selected, the group determination is completed, and when unselected resources remain, the process returns to step S20.

FIG. 15 is a flowchart showing a procedure example of the virtual machine start process.
(S30) The communication log acquisition unit 223 acquires the communication log during the test execution. The communication log includes the destination and the transmission time of the message transmitted by the virtual machine 51.

(S31) The communication log acquisition unit 223 transmits the communication log to the information collection server 100.
(S32) The operation information acquisition unit 224 receives a collection frequency notification from the information collection server 100. The collection frequency notification indicates a predetermined low collection frequency as an initial value of the collection frequency.

(S33) The operation information acquisition unit 224 starts acquiring operation information of all resources and transmitting the operation information to the information collection server 100 at the low collection frequency notified in step S32.
(S34) The group control unit 225 determines whether or not the group notification has been received from the information collection server 100. The group notification includes the identification information of the subordinate virtual machines belonging to the same group and the magnification thereof. If the group notification is received, the process proceeds to step S35, and if the group notification is not received, the process proceeds to step S38.

(S35) The group control unit 225 calculates the collection target ratio of the virtual machines in the group based on the group notification. Assuming that the total number of virtual machines belonging to the group is n and the number of dependent virtual machines that do not collect operation information is m, the collection target ratio is (nm) ÷ n.

(S36) The group control unit 225 raises the collection frequency of the operation information acquisition unit 224 from the initial value notified in step S32 according to the collection target ratio calculated in step S35. The smaller the collection target ratio, the higher the collection frequency. The changed collection frequency is the initial value multiplied by n ÷ (nm) × α. Therefore, the transmission interval after the change is shortened to the initial transmission interval multiplied by (nm) ÷ n ÷ α. However, if the collection frequency calculated by this formula exceeds a predetermined upper limit, the upper limit shall be the collection frequency.

(S37) The group control unit 225 transmits a stop notification to the subordinate virtual machine.
(S38) The operation information acquisition unit 224 determines whether or not the stop notification has been received from the representative virtual machine. If the stop notification is received, the process proceeds to step S39, and if the stop notification is not received, the virtual machine start process ends. The representative virtual machine and the subordinate virtual machine are classified for each resource. Therefore, the virtual machine 51 may be designated as a representative virtual machine for a certain resource and as a subordinate virtual machine for another resource. In that case, the virtual machine 51 may receive a group notification from the information collection server 100 for a certain resource and a stop notification from the representative virtual machine for another resource.

(S39) The operation information acquisition unit 224 stops the acquisition and transmission of the operation information.
FIG. 16 is a flowchart showing a procedure example of virtual machine continuous processing.
(S40) The group control unit 225 determines whether the virtual machine 51 is a representative virtual machine. If it is a representative virtual machine, the process proceeds to step S41, and if it is not a representative virtual machine, the process proceeds to step S50. Whether or not it is a representative virtual machine is determined for each resource.

(S41) The group control unit 225 identifies a subordinate virtual machine that belongs to the same group as the virtual machine 51 with reference to the group management table 231. The group control unit 225 requests the operation information from the specified subordinate virtual machine and receives the operation information.

(S42) The group control unit 225 records the measured value included in the operation information of the dependent virtual machine received in step S41 in association with the latest measured value of the virtual machine 51 acquired by the operation information acquisition unit 224. To do. The group control unit 225 calculates the correlation coefficient between the measured value of the virtual machine 51 and the measured value of the dependent virtual machine based on the record of the latest fixed period.

(S43) The group control unit 225 compares the absolute value of the correlation coefficient calculated in step S42 with a predetermined threshold value. If the absolute value of the correlation coefficient is greater than the threshold value, the process proceeds to step S44, and if the absolute value of the correlation coefficient is equal to or less than the threshold value, the process proceeds to step S46.

(S44) The group control unit 225 calculates the magnification of the measured value of the dependent virtual machine with respect to the measured value of the virtual machine 51 based on the record of the latest fixed period. The group control unit 225 updates the magnification recorded in the group management table 231.

(S45) The group control unit 225 transmits a correlation update notification including the magnification updated in step S44 to the information collection server 100. Then, the process proceeds to step S50.
(S46) The group control unit 225 transmits a restart notification to the subordinate virtual machine.

(S47) The group control unit 225 updates the collection target ratio of the virtual machines in the group when one subordinate virtual machine leaves the group. Assuming that the total number of virtual machines belonging to the group at the immediately preceding point in time is n and the number of dependent virtual machines that do not collect operation information at the immediately preceding point in time is m, the collection target ratio increases to (nm) ÷ (n-1). To do.

(S48) The group control unit 225 lowers the collection frequency of the operation information acquisition unit 224 from the immediately preceding value according to the collection target ratio calculated in step S47.
(S49) The group control unit 225 transmits a group release notification including the identification information of the virtual machine out of the group to the information collection server 100.

(S50) The operation information acquisition unit 224 determines whether or not the restart notification has been received from the representative virtual machine. If the restart notification is received, the process proceeds to step S51, and if the restart notification is not received, the virtual machine continuation process ends. The virtual machine 51 may be designated as a representative virtual machine for a certain resource and as a subordinate virtual machine for another resource. Therefore, the virtual machine 51 may monitor the dependent virtual machine for one resource while receiving a restart notification from the representative virtual machine for another resource.

(S51) The operation information acquisition unit 224 resumes acquisition and transmission of operation information. The collection frequency after resumption is, for example, the low frequency initial value notified in step S32.
FIG. 17 is a flowchart showing an example of a procedure for continuous server processing.

(S60) The correlation update unit 127 receives the correlation update notification from each representative virtual machine. The correlation update notification includes the identification information of the dependent virtual machine and the updated magnification.
(S61) The correlation update unit 127 updates the magnification of the correlation table 133.

(S62) The correlation update unit 127 determines whether the group release notification has been received from any of the representative virtual machines. If the group release notification is received, the process proceeds to step S63, and if the group release notification is not received, the process proceeds to step S64.

(S63) The correlation update unit 127 deletes the correlation between the representative virtual machine that sent the group release notification and the notified dependent virtual machine from the correlation table 133.
(S64) The correlation detection unit 126 redetermines whether or not any of the current representative virtual machines can be grouped according to the flowchart shown in FIG.

(S65) The correlation detection unit 126 determines whether or not the number of subordinate virtual machines has increased due to the change of the representative virtual machine to the subordinate virtual machine by the redetermination in step S63. If the number of dependent virtual machines has increased, the process proceeds to step S66, and if the number has not increased, the process proceeds to step S67.

(S66) The correlation detection unit 126 updates the correlation table 133. In addition, the correlation detection unit 126 transmits a group notification including the identification information of the subordinate virtual machine and its magnification to the representative virtual machine of the group to which the new subordinate virtual machine belongs.

(S67) The operation status display unit 128 determines whether the request for the operation status has been received from the user. If the request for the operating status is accepted, the process proceeds to step S68, and if the request for the operating status is not accepted, the server continuation process ends.

(S68) The operation status display unit 128 estimates a numerical value indicating the operation status of the dependent virtual machine by multiplying the measured value of the representative virtual machine registered in the correlation table 133 by a multiplication factor.
(S69) The operation status display unit 128 generates an operation status table 134 that lists the numerical values of the operation status of each virtual machine, and displays the operation status table 134 on the display device 111. The numerical values in the operation status table 134 include the collected measured values and the estimated values calculated in step S68.

According to the information processing system of the second embodiment, the collection of operation information from some virtual machines is stopped, and the collection of operation information from the remaining virtual machines is collected within the range in which the total data communication volume is reduced. The frequency will be increased. For virtual machines that do not collect operation information, resource usage is estimated based on the correlation with the virtual machines that collect operation information. As a result, it is possible to ensure real-time monitoring of virtual machines and improve monitoring accuracy while reducing network traffic. In particular, in a multi-cloud environment, it is possible to reduce the traffic of a wide area data communication network such as the Internet. Therefore, it is possible to reduce the monitoring cost of the virtual machine and suppress the delay in monitoring.

In addition, the test request is used to determine the data flow between a plurality of virtual machines in the service, and the correlation between the parent virtual machine and the child virtual machine is determined along the data flow. Since the service sends messages from the parent virtual machine to the child virtual machine, there is likely a strong correlation between the load on the parent virtual machine and the load on the child virtual machine. Therefore, the correlation can be efficiently determined among a plurality of virtual machines. Also, changes in the correlation between the representative virtual machine and the dependent virtual machines are distributed and monitored by the representative virtual machines. Therefore, the traffic of the wide area data communication network can be reduced.

10, 20, 20a Information processing device 11 Communication unit 12 Storage unit 13 Processing unit 14 Correlation 21,22 Virtual machine 23,24 Operation information

Claims (13)

A second operation information indicating the resource usage status in the first virtual machine is received from the first virtual machine, and a second operation information indicating the resource usage status in the second virtual machine is received from the second virtual machine. Communication unit that receives operation information and
A storage unit that stores the received history of the first operation information and the second operation information, and
Based on the history, the correlation of the resource usage status between the first virtual machine and the second virtual machine is determined, and when the correlation satisfies a predetermined condition, the second operation information Is controlled so that the reception frequency of the first operation information is lower than the reception frequency of the first operation information, and based on the first operation information received from the first virtual machine and the correlation, the second operation information A processing unit that estimates resource usage in a virtual machine,
Information processing device with. When the correlation satisfies the predetermined condition, the transmission of the second operation information from the second virtual machine to the information processing device is controlled to be stopped.
The information processing device according to claim 1. When the correlation satisfies the predetermined condition, it is instructed to reduce the frequency of transmitting the second operation information to the second virtual machine through the first virtual machine.
The information processing device according to claim 1. Before determining the correlation, the processing unit receives the first operation information and the first operation information at a predetermined frequency lower than the reception frequency of the first operation information when the correlation satisfies the predetermined condition. Control so that the operation information of 2 is received,
The information processing device according to claim 1. When the correlation satisfies the predetermined condition, the reception frequency of the second operation information is lower than that before the correlation is determined, and the reception frequency of the first operation information is before the correlation is determined. Controlled to be higher than
The information processing device according to claim 1. Of the plurality of virtual machines including the first virtual machine and the second virtual machine, the frequency of receiving operation information of some of the virtual machines including the second virtual machine is higher than before the correlation is determined. Also, the frequency of receiving operation information of other virtual machines including the first virtual machine becomes higher than before the correlation is determined.
The reception frequency after the change of the first operation information is determined based on the ratio of the number of the other virtual machines to the number of the plurality of virtual machines.
The information processing device according to claim 5. The reception frequency of the second operation information is updated according to the change in the correlation.
The information processing device according to claim 1. The change in the correlation is monitored by the first virtual machine, and the first virtual machine notifies the information processing apparatus.
The information processing device according to claim 7. The processing unit detects the order of message communication between the first virtual machine and a plurality of virtual machines including the second virtual machine by sending a test request, and the virtual machine from which the message is sent. Is selected as the first virtual machine, and the virtual machine to which the message is sent is selected as the second virtual machine.
The information processing device according to claim 1. Based on the first operation information and the second operation information within a predetermined time included in the history, the processing unit has a time series of resource usage status in the first virtual machine as the correlation. A correlation coefficient between the change and the time-series change of the resource usage in the second virtual machine is calculated, and it is determined whether the correlation coefficient is larger than the threshold value.
The information processing device according to claim 1. The first information processing device that executes the first virtual machine,
A second information processing device that executes the second virtual machine,
A third information processing device that communicates with the first virtual machine and the second virtual machine,
The third information processing apparatus has
The first operation information indicating the resource usage status in the first virtual machine is received from the first virtual machine, and the resource usage status in the second virtual machine is shown from the second virtual machine. Received the operation information of 2 and
The received history of the first operation information and the second operation information is retained, and the history is retained.
Based on the history, the correlation of the resource usage status between the first virtual machine and the second virtual machine is determined, and when the correlation satisfies a predetermined condition, the second operation information Is controlled so that the reception frequency of the first operation information is lower than the reception frequency of the first operation information.
Based on the first operation information received from the first virtual machine and the correlation, the resource usage status in the second virtual machine is estimated.
Information processing system. The computer
The first operation information indicating the resource usage status in the first virtual machine received from the first virtual machine and the resource usage status in the second virtual machine received from the second virtual machine are shown. Acquire the history with the second operation information shown,
Based on the history, the correlation of the resource usage status between the first virtual machine and the second virtual machine is determined, and when the correlation satisfies a predetermined condition, the second operation information Is controlled so that the reception frequency of the first operation information is lower than the reception frequency of the first operation information.
Based on the first operation information received from the first virtual machine and the correlation, the resource usage status in the second virtual machine is estimated.
Information processing method. On the computer
The first operation information indicating the resource usage status in the first virtual machine received from the first virtual machine and the resource usage status in the second virtual machine received from the second virtual machine are shown. Acquire the history with the second operation information shown,
Based on the history, the correlation of the resource usage status between the first virtual machine and the second virtual machine is determined, and when the correlation satisfies a predetermined condition, the second operation information Is controlled so that the reception frequency of the first operation information is lower than the reception frequency of the first operation information.
Based on the first operation information received from the first virtual machine and the correlation, the resource usage status in the second virtual machine is estimated.
A program that executes processing.

Images