JP6924072B2 - Management server, cloud system, and management method - Google Patents

Description

The present invention relates to a technique for managing a cloud system.

Conventionally, there is a technology for detecting a failure or a load state of a CPU and excluding a server or an application from the load balancing target of a cloud system according to the detection result. As a method of detecting the load state of the CPU, for example, Patent Document 1 discloses a method of efficiently measuring and collecting different server performance for each application.

Japanese Unexamined Patent Publication No. 2006-285316

However, if the server or application is operating with a delay, it does not mean that the server or application is out of order, so the failure cannot be detected and the server or application that is operating with a delay cannot be detected. Could not be excluded from the load balancing target of the cloud system. As a result, processing delays (defects) may occur in the services provided by the cloud system.

Further, even if the CPU is in a high load state, if there is no problem in the service provided by the cloud system, it is not necessary to exclude it from the target of load balancing. Nevertheless, until now, servers with a high CPU load have been uniformly excluded from the load balancing target of the cloud system, so there is a risk that the server will be excluded from the load balancing target more than necessary. This is because it was not possible to determine whether the service provided by the cloud system has a problem due to the high load state of the CPU.

In view of the above problems, an object of the present invention is to provide a management technique for appropriately suppressing the occurrence of a defect in a service provided by a cloud system.

The management server according to the present invention is a management server that can be incorporated into a cloud system including a plurality of managed servers and a load balancer that distributes a processing load to the plurality of managed servers. When the test signal is transmitted, the transmitter that transmits the test signal to the test communication terminal via each of the applications executed by the managed server and the response signal to the test signal transmitted from the test communication terminal are transmitted. The test signal when the receiving unit receives via the used application and the receiving unit does not receive the response signal even after a predetermined time has elapsed since the transmitting unit transmitted the test signal. Exclude the application used at the time of transmission from the distribution target of the processing load alone or together with other applications belonging to the same virtual server, or together with other applications belonging to the same physical server. It is a configuration (first configuration) including an instruction unit that outputs an instruction signal to the load balancer.

In the management server of the first configuration, the application excluded from the distribution target of the processing load is a single application, or together with other applications belonging to the same virtual server, or other application belonging to the same physical server. After restarting with the application, the transmitting unit transmits the test signal to the test communication terminal via each of the restarted applications, and the receiving unit transmits the restarted application to the test communication terminal. When the receiving unit receives the response signal before a predetermined time elapses after the transmitting unit transmits the test signal, the indicating unit issues a release signal for canceling the instruction content of the instruction signal. It may be a configuration (second configuration) that outputs to the load balancer.

The management server having the first or second configuration may further include an operation unit that operates the transmission timing of the transmission unit (third configuration).

In the management server having any of the first to third configurations, the transmission unit may transmit the test signal to a plurality of test communication terminals existing at different locations (fourth configuration). good.

In the management server of the fourth configuration, the cloud system further includes a plurality of client terminals connected to the plurality of managed servers via a network via the load balancer, and the plurality of client terminals move differently. It may be a configuration (fifth configuration) that is mounted on the body and also functions as the test communication terminal.

In the management server having the fourth or fifth configuration, the calculation unit calculates the average value of the time required from the transmission unit transmitting the test signal to the reception unit receiving the response signal for each location. It may be a configuration (sixth configuration) further comprising.

In the management server having any of the first to sixth configurations, the application may have a configuration for processing an audio signal (seventh configuration).

In the management server having any of the first to seventh configurations, the instruction signal is a signal that excludes the application from the distribution target of the processing load, and the transmission unit is the test signal for all the applications. If the receiving unit does not receive the response signal even after a predetermined time has elapsed since the transmission of the above, or the instruction signal is a signal that excludes the application from the distribution target of the processing load for each virtual server. When the receiving unit does not receive the response signal even after a predetermined time has elapsed since the transmitting unit transmits the test signal for at least one application in each of the virtual servers, or the instruction signal is transmitted. A signal for excluding the application from the distribution target of the processing load for each physical server, and a predetermined time has elapsed since the transmission unit transmitted the test signal for at least one application in each of the physical servers. However, when the receiving unit does not receive the response signal, the indicating unit may be exceptionally configured not to output the instruction signal (eighth configuration).

The cloud system according to the present invention includes a plurality of managed servers, a load balancer that distributes a processing load to the plurality of managed servers, and a management server having any of the first to eighth configurations. It is a configuration to be provided (a ninth configuration).

In the cloud system having the ninth configuration, the two management servers are provided, the plurality of managed servers, the load balancer, and one of the management servers are installed in the data center, and the data center and the other are installed. The configuration may be such that the management server is connected to the network (tenth configuration).

In the cloud system having the tenth configuration, the instruction signal is a signal that excludes the application from the distribution target of the processing load by itself, and the transmission unit performs the test for all the applications in the tests by both management servers. When the receiving unit does not receive the response signal even after a predetermined time has elapsed after transmitting the test signal, or when the instruction signal excludes the application from the distribution target of the processing load in units of the virtual server. It is a signal, and in a test by both management servers, even if a predetermined time elapses after the transmission unit transmits the test signal for at least one application in each of the virtual servers, the reception unit still receives the response signal. Is not received, or the instruction signal is a signal that excludes the application from the distribution target of the processing load in the physical server unit, and at least one of the above in each of the physical servers in a test by both management servers. When the receiving unit does not receive the response signal even after a predetermined time has elapsed after the transmitting unit transmits the test signal with respect to the application, the indicating unit exceptionally does not output the instruction signal, and a part of the instruction signal is not output. The processing priority of the application may be higher than the processing priority of the remaining applications (11th configuration).

The management method according to the present invention is a management method of a cloud system including a plurality of managed servers and a load balancer that distributes a processing load to the plurality of managed servers, and the plurality of managed servers. A transmission step of transmitting a test signal to the test communication terminal via each application executed by the server and a response signal to the test signal transmitted from the test communication terminal were used at the time of transmitting the test signal. When the reception process receives via the application and the reception process does not receive the response signal even after a predetermined time has elapsed after the transmission process transmits the test signal, when the test signal is transmitted. An instruction signal for excluding the used application from the distribution target of the processing load alone or together with other applications belonging to the same virtual server, or together with other applications belonging to the same physical server. It is a configuration (12th configuration) including an instruction step for outputting to the load balancer.

According to the management technique of the present invention, it is possible to appropriately suppress the occurrence of a defect in the service provided by the cloud system.

Diagram showing a configuration example of a cloud system The figure explaining the operation when the vehicle allocation instruction is transmitted Diagram showing a configuration example of a monitoring server The figure explaining the operation when a test signal is transmitted Flowchart showing an operation example of the management server Flowchart showing other operation examples of the management server Diagram showing other configuration examples of the monitoring server Diagram showing a configuration example of a cloud system

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the drawings.

<1. Cloud system configuration>
FIG. 1 is a diagram showing a configuration example of a cloud system. The cloud system shown in FIG. 1 includes a management server 300 and a data center 700. Hereinafter, a case where the cloud system shown in FIG. 1 is used as a taxi dispatch system will be described.

The client terminal 100 and the test communication terminal 400 can communicate with the data center 700 via the public network 500.

The vehicle allocation center 200 and the management server 300 can communicate with the data center 700 via the dedicated network 600. In addition, unlike this embodiment, the vehicle allocation center 200 and the management server 300 may be able to communicate with the data center 700 via the public network 500.

The client terminal 100 is permanently or temporarily mounted on the taxi. Although only one client terminal 100 is shown in FIG. 1, there are actually as many client terminals 100 as the number of taxis using the cloud system shown in FIG. The client terminal 100 includes a handheld microphone 101, an information processing device 102, a position information acquisition unit 103, and an IP radio 104.

The handheld microphone 101 converts the vehicle allocation instruction, which is an electric signal, into voice and transmits it to the driver, converts the response issued by the driver into an instruction response, which is an electric signal, and outputs the response to the information processing device 102.

The information processing device 102 processes a vehicle allocation instruction, an instruction response, position information, and the like.

The position information acquisition unit 103 acquires the position information of the client terminal 100 and thus the position information of the taxi on which the client terminal 100 is mounted. As the position information acquisition unit 103, for example, a GPS (Global Positioning System) receiving unit used in a navigation system or the like can be used. The GPS receiving unit receives signals from a plurality of GPS satellites and acquires information in which the position of the client terminal 100 at the current time is expressed in longitude and latitude on the earth.

The IP radio 104 communicates with the M2M communication device 701 of the data center 700 via the public network 500.

The vehicle dispatch center 200 includes personal computers (hereinafter abbreviated as “PC”) 201 and 202, a switch 203, a local server 204, and a router 205. Although only one vehicle allocation center 200 is shown in FIG. 1, there are actually as many vehicle allocation centers 200 as there are taxi companies that use the cloud system shown in FIG.

The switch 203 switches the PC connected to the local server 204. The local server 204 transmits a vehicle allocation instruction to the data center 700 via the router 205 and the dedicated network 600 in response to a command from the PC 201 or 202.

The management server 300 includes a PC 301, a monitoring server 302, and a router 303. The monitoring server 302 periodically transmits a test signal to the data center 700 via the router 303 and the dedicated network 600. Further, the monitoring server 302 may transmit a test signal to the data center 700 via the router 303 and the dedicated network 600 in response to a command from the PC 301.

The test communication terminal 400 includes an information processing device 401 and an IP wireless communication device 402. When the information processing device 401 receives the test signal, it outputs a response signal.

The IP radio 402 communicates with the M2M communication device 701 of the data center 700 via the public network 500.

The data center 700 includes an M2M communication machine 701, a router 702, a load balancer 703, switches 704 and 707, physical servers 705 and 706, database servers 708 and 709, and a database 710.

The data center 700 uses the M2M communication device 701 and the router 702 to communicate with the client terminal 100, the vehicle allocation center 200, the management server 300, and the test communication terminal 400.

The load balancer 703 distributes the processing load to the physical servers 705 and 706 and the virtual servers included in each physical server.

The switch 704 switches the physical server connected to the load balancer 703.

The physical server 705 includes virtual servers 705a and 705b. The virtual server 705a includes applications 705a1 to 705a3, a voice control unit 705a4, and a storage unit 705a5. The virtual server 705a may be configured to include, for example, four CPUs, and each CPU processes any one of the applications 705a1 to 705a3 and the voice control unit 705a4. It may be. The voice control unit 705a4 cooperates with each of the applications 705a1 to 705a3, and the storage unit 705a5 stores the processing contents of each of the applications 705a1 to 705a3 as log data.

The virtual server 705b has the same configuration as the virtual server 705a, and the physical server 706 has the same configuration as the physical server 705. The virtual server and the physical server described above are managed servers managed by the management server 300. If there are a plurality of managed servers, the number of virtual servers and the number of physical servers do not have to be the same as in the present embodiment.

The switch 707 switches the database server connected to the load balancer 703.

The database servers 708 and 709 update the database 710 based on the data received by the data center 700 and the like. For example, the location information of each client terminal 100 stored in the database 710 is updated.

<2. Operation when a vehicle dispatch instruction is sent>
FIG. 2 is a diagram illustrating an operation when a vehicle allocation instruction is transmitted. When the PC 201 or 202 transmits the vehicle allocation instruction to the local server 204, the local server 204 transmits the vehicle allocation instruction to a certain application of the physical server 705 or 706. One application of the physical server 705 or 706 is designated by the load balancer 703.

One application of the physical server 705 or 706 confirms the destination in the database 710 based on the information about the vehicle allocation included in the vehicle allocation instruction. The database 710 instructs one application of the physical server 705 or 706 to send a predetermined number of client terminals 100 located in the vicinity of the vehicle allocation destination.

One application of the physical server 705 or 706 transmits the vehicle allocation instruction to the information processing device 102 of a predetermined number of client terminals 100 located in the vicinity of the vehicle allocation destination according to the transmission destination instruction from the database 710.

One application of the physical server 705 or 706 receives the earliest returned vehicle allocation instruction response from the information processing device 102 that has received the vehicle allocation instruction, and transmits the received vehicle allocation instruction response to the local server 204. As a result, one vehicle allocation is completed.

<3. Operation when a test signal is transmitted>
Before explaining the operation when the test signal is transmitted, the configuration of the monitoring server 302 that transmits the test signal will be described. FIG. 3 is a diagram showing a configuration example of the monitoring server.

The monitoring server 302 shown in FIG. 3 includes a control unit 302a and a storage unit 302b. The control unit 302a is a microcomputer including a CPU, RAM, and ROM. The control unit 302a includes a transmission unit 302a1 that transmits a test signal, a reception unit 302a2 that receives a response signal to the test signal, and an instruction unit 302a3 that outputs an instruction signal to the load balancer 703. The storage unit 302b is a non-volatile storage medium such as an EEPROM, a flash memory, or a hard disk drive. The storage unit 302b includes a program 302b1 and managed server data 302b2. The program 302b1 is read by the control unit 302a and executed for the control unit 302a to control the monitoring server 302. The managed server data 302b2 is, for example, data indicating the relationship between each managed server and each application, and is used when the transmission unit 302a1 transmits a test signal.

FIG. 4 is a diagram illustrating an operation when a test signal is transmitted. The monitoring server 302 sends a test signal to one application on the physical server 705 or 706. One application of the physical server 705 or 706 is designated by the monitoring server 302.

One application on the physical server 705 or 706 confirms the destination in the database 710 based on the test signal. The database 710 instructs one application of the physical server 705 or 706 to send to the information processing device 401 of the test communication terminal 400.

One application of the physical server 705 or 706 transmits a test signal to the information processing device 401 of the test communication terminal 400 according to a destination instruction from the database 710.

One application of the physical server 705 or 706 receives the response signal returned from the information processing apparatus 401 that has received the test signal, and transmits the received response signal to the monitoring server 302. This completes one test of one application on the physical server 705 or 706.

<4. Operation of monitoring server>
FIG. 5 is a flowchart showing an operation example of the management server 300. The management server 300 periodically (for example, every few seconds) starts the flow operation shown in FIG. First, the transmission unit 302a1 of the monitoring server 302 transmits a test signal to the test communication terminal 400 via a certain application of the physical server 705 or 706 (step S11). Next, the control unit 302a of the monitoring server 302 used the response signal to the test signal at the time of transmitting the test signal even after a predetermined time has elapsed since the transmission unit 302a1 of the monitoring server 302 transmitted the test signal. It is determined whether or not the reception unit 302a2 of the monitoring server 302 has received the signal via the application (step S12).

When the reception unit 302a2 of the monitoring server 302 does not receive the response signal within a predetermined time, the instruction unit 302a3 of the monitoring server 302 load balances the instruction signal that excludes the application used at the time of transmitting the test signal from the distribution target. The output is output to the device 703 (step S13), and the flow operation shown in FIG. 5 is terminated. As a method of excluding from the distribution target, (I) a method of excluding the application used at the time of transmitting the test signal by itself, and (II) a method of excluding the application used at the time of transmitting the test signal belong to the same virtual server. Either the method of excluding the application together with other applications or the method of excluding the application used at the time of transmitting the test signal together with other applications belonging to the same physical server may be adopted. For example, when excluding application 705a1 from the distribution target, application 705a1 may be excluded alone, applications 705a1 to 705a3 belonging to the same virtual server 705a may be excluded, and all applications belonging to the same physical server 705 may be excluded. You may exclude all of the applications.

On the other hand, when the receiving unit 302a2 of the monitoring server 302 receives the response signal within a predetermined time, the instruction unit 302a of the monitoring server 302 ends the flow operation shown in FIG. 5 without outputting the instruction signal.

The operation flow shown in FIG. 5 is executed by changing the application used at the time of transmitting the test signal every time, and the test is executed for all the applications. However, for the applications excluded from the distribution target, the test is not executed until the exclusion of the distribution target is released.

According to the test executed by the management server 300, when the processing delay actually occurs due to the test, the application used at the time of transmitting the test signal is excluded from the distribution target, so that exclusion omission or excessive exclusion is prevented. be able to. Therefore, it is possible to appropriately suppress the occurrence of a defect in the service provided by the cloud system. In particular, taxi dispatch systems are required to continue to operate normally 24 hours a day, 365 days a year, so it is important to eliminate processing delays and prevent problems with the services provided by the cloud system. be. Further, in the taxi dispatch system, since the application processes the voice signal, if the processing is delayed, the voice that the listener cannot recognize is output, which is a serious obstacle. Therefore, if the application is processing audio signals, the tests performed by management server 300 are particularly useful.

In the above-described embodiment, the management server 300 periodically starts the flow operation shown in FIG. 5, but instead of or in addition to the flow operation, an operation unit such as a keyboard or a pointing device of the PC 301 sets the transmission timing of the transmission unit 302a1. You may try to operate it. As a result, for example, when the operator of the vehicle allocation center 200 complains to the operator of the management server 300 regarding a problem with the service provided by the cloud system, the cause can be immediately investigated.

FIG. 6 is a flowchart showing another operation example of the management server 300. The management server 300 starts the flow operation shown in FIG. 6 after a single application excluded from the processing load distribution target is restarted. The restart of the application may be executed by an instruction from the management server 300, or may be executed autonomously by the application itself.

First, the transmission unit 302a1 of the monitoring server 302 transmits a test signal to the test communication terminal 400 via the restarted application (step S21). Next, the control unit 302a of the monitoring server 302 used the response signal to the test signal at the time of transmitting the test signal even after a predetermined time has elapsed since the transmission unit 302a1 of the monitoring server 302 transmitted the test signal. It is determined whether or not the reception unit 302a2 of the monitoring server 302 has received the signal via the application (step S22).

When the receiving unit 302a2 of the monitoring server 302 receives the response signal within a predetermined time, the instruction unit 302a3 of the monitoring server 302 releases the instruction content of the instruction signal to the application used at the time of transmitting the test signal. Is output to the load balancer 703 (step S23), and the flow operation shown in FIG. 6 is terminated. In the flow operation shown in FIG. 5, when the application used at the time of transmitting the test signal is excluded together with other applications belonging to the same virtual server, the application belonging to the same virtual server excluded from the distribution target. Restart. Then, if the receiving unit 302a2 of the monitoring server 302 receives the response signal within the predetermined time for all the restarted applications, the indicating unit 302a3 of the monitoring server 302 outputs the release signal to the load balancer 703. To. Similarly, in the flow operation shown in FIG. 5, when the application used at the time of transmitting the test signal is excluded together with other applications belonging to the same physical server, it belongs to the same physical server excluded from the distribution target. Restart the application. Then, if the receiving unit 302a2 of the monitoring server 302 receives the response signal within the predetermined time for all the restarted applications, the indicating unit 302a3 of the monitoring server 302 outputs the release signal to the load balancer 703. To.

On the other hand, when the receiving unit 302a2 of the monitoring server 302 does not receive the response signal within the predetermined time, the indicating unit 302a of the monitoring server 302 ends the flow operation shown in FIG. 6 without outputting the release signal.

By executing the above-mentioned de-exclusion from the distribution target, the application whose processing delay has been eliminated can be smoothly returned to the distribution target.

<5. Modification example>
It should be considered that the embodiments are exemplary in all respects and are not restrictive, and the technical scope of the invention is indicated by the claims rather than the description of the embodiments. It should be understood that it is intended and includes all changes that fall within the meaning and scope of the claims. The various modifications described in the above-described embodiments can be appropriately combined and carried out as long as they do not contradict each other.

For example, in the above-described embodiment, the management server 300 transmits a test signal to one test communication terminal 400, but the management server 300 sends a test signal to a plurality of test communication terminals 400 existing in different locations. You may send it.

Is the cause of the problem in the service provided by the cloud system caused by the management server 300 transmitting the test signal to the plurality of test communication terminals 400 existing in different locations on the data center 700 side? It is possible to distinguish whether it is on the public network 500 side.

In the test for one application of the physical server 705 or 706, if the processing is delayed for all of the plurality of test communication terminals 400, the cause of the problem in the service provided by the cloud system is on the data center 700 side. Can be estimated. On the other hand, when a processing delay occurs in only a part of a plurality of test communication terminals 400 in a test for a certain application of the physical server 705 or 706, the cause of a malfunction in the service provided by the cloud system is. It can be estimated that it is on the client terminal 100 or the public network 500 side corresponding to the location of the test communication terminal 400 where the processing delay has occurred. Therefore, the control unit 302a of the monitoring server 302 may execute the estimation process as described above.

For example, by adding the function of the information processing device 401 of the test communication terminal 400 to the information processing device 102 of the client terminal 100, the plurality of client terminals 100 can be used as a plurality of test communication terminals 400 existing in different places. be able to. As a result, the cost required for installing the test communication terminal 400 can be suppressed. Further, since the location where the terminal exists in the test and the location where the terminal exists when the service is actually used match, the estimation accuracy in the above-mentioned estimation process is improved.

Further, when the management server 300 transmits a test signal to a plurality of test communication terminals 400 existing at different locations, the monitoring server 302 may be configured as shown in FIG. 7. In the monitoring server 302 shown in FIG. 7, the control unit 302a includes a calculation unit 302a4.

The calculation unit 302a4 calculates the average value of the time required from the transmission unit 302a1 transmitting the test signal to the receiving unit 302a2 receiving the response signal for each location. By confirming the average time calculated by the calculation unit 302a4, it is possible to easily confirm whether or not the processing delay is a problem in a specific area.

The average value may be an average value for all applications or an average value for a specific application. Further, for example, the average value of the time belonging to the central range may be calculated by omitting a predetermined number of hours from the larger one and omitting a predetermined number of hours from the smaller one. Further, each place may be one point on the map, but a predetermined group of areas may be the same place. The average time calculated by the calculation unit 302a4 may be mapped on a map, for example, and the mapping result may be displayed on the display of the PC 301. This makes it even easier to see if processing delays are a problem in a particular region. The area is, for example, a range (boundary range) determined according to the boundaries of prefectures and municipalities. In addition, the area may be the respective range (business range) defined as the area where each taxi company can operate. Further, the area may be a range (communication network range) determined according to the type of communication network (for example, FORMA communication network, wired / wireless communication network, etc.).

Further, if the above-described embodiment is implemented as it is, all the applications may be excluded from the distribution target, and a situation may occur in which the cloud system does not operate at all. Therefore, if all the applications are excluded from the distribution target if the instruction signal is output as it is, the instruction unit 302a3 of the monitoring server 302 is configured to exceptionally not output the instruction signal, and all the applications are the distribution target. You may try to avoid the situation of being excluded from.

Further, for example, in the above-described embodiment, only one management server 300 is provided, but a management server 800 may be provided inside the data center 700 as shown in FIG. 8 in addition to the management server 300.

By providing management servers inside and outside the data center 700, it is possible to determine whether the cause of the trouble in the service provided by the cloud system is on the data center 700 side or the dedicated network 600 side.

In the test for one application of the physical server 705 or 706, if the processing delay occurs for both management servers, it can be estimated that the cause of the failure in the service provided by the cloud system is on the data center 700 side. On the other hand, in the test for a certain application of the physical server 705 or 706, if the processing delay occurs only for the management server 300, the cause of the trouble in the service provided by the cloud system is on the dedicated network 600 side. Can be estimated. Therefore, the management server 300 or the management server 800 may acquire the test result on the other management server and execute the estimation process as described above.

Further, in the test for one application of the physical server 705 or 706, if the processing is delayed for both management servers and the instruction signal is output as it is, all the applications are excluded from the distribution target. If this happens, both management servers will exceptionally not output an instruction signal to avoid the situation where all applications are excluded from the distribution target, and the processing of some applications is prioritized. The degree (for example, NICE value) may be higher than the processing priority of the remaining applications. By using an application with a high processing priority, processing delay can be suppressed. In this case, the CPU of the virtual server may be a multi-core CPU.

300 Management Server 301 PC
302a1 Transmitter 302a2 Receiver 302a3 Indicator 302a4 Calculation unit 500 Public network 600 Dedicated network 700 Data center 703 Load balancer 705, 706 Physical server 705a, 705b Virtual server

Claims (12)

A management server that can be incorporated into a cloud system including a plurality of managed servers and a load balancer that distributes a processing load to the plurality of managed servers.
A transmitter that transmits a test signal to the test communication terminal via each of the applications executed by the plurality of managed servers, and a transmitter.
A receiving unit that receives a response signal to the test signal transmitted from the test communication terminal via the application used at the time of transmitting the test signal.
When the receiving unit does not receive the response signal even after a predetermined time has elapsed after the transmitting unit transmits the test signal, the application used at the time of transmitting the test signal may be used alone or. An instruction unit that outputs an instruction signal to be excluded from the distribution target of the processing load to the load balancer together with the other application belonging to the same virtual server or together with the other application belonging to the same physical server.
A management server. After the application excluded from the distribution target of the processing load is restarted alone, with other applications belonging to the same virtual server, or with other applications belonging to the same physical server.
The transmitting unit transmits the test signal to the test communication terminal via each of the restarted applications, and the receiving unit transmits the test signal to the test communication terminal, and the transmitting unit transmits the test signal to all the restarted applications. When the receiving unit receives the response signal before the predetermined time elapses,
The management server according to claim 1, wherein the instruction unit outputs a release signal for releasing the instruction content of the instruction signal to the load balancer. The management server according to claim 1 or 2, further comprising an operation unit that operates the transmission timing of the transmission unit. The management server according to any one of claims 1 to 3, wherein the transmission unit transmits the test signal to a plurality of test communication terminals existing at different locations. The cloud system further includes a plurality of client terminals connected to the plurality of managed servers via the load balancer.
The management server according to claim 4, wherein the plurality of client terminals are mounted on different mobile bodies and also function as the test communication terminal. The fourth or fifth aspect of the present invention further comprises a calculation unit that calculates an average value of the time required for the reception unit to receive the response signal after the transmission unit transmits the test signal for each location. Management server. The management server according to any one of claims 1 to 6, wherein the application processes an audio signal. The instruction signal is a signal that excludes the application from the distribution target of the processing load by itself, and the receiving unit even if a predetermined time elapses after the transmitting unit transmits the test signal for all the applications. Does not receive the response signal, or
The instruction signal is a signal for excluding the application from the distribution target of the processing load for each virtual server, and is determined after the transmission unit transmits the test signal for at least one application in each of the virtual servers. When the receiving unit does not receive the response signal even after a lapse of time, or
The instruction signal is a signal that excludes the application from the distribution target of the processing load for each physical server, and is determined after the transmission unit transmits the test signal for at least one application in each of the physical servers. If the receiver does not receive the response signal over time,
The management server according to any one of claims 1 to 7, wherein the instruction unit exceptionally does not output the instruction signal. With multiple managed servers
A load balancer that distributes the processing load to the plurality of managed servers, and
The management server according to any one of claims 1 to 8.
A cloud system equipped with. Equipped with two management servers
The cloud system according to claim 9, wherein the plurality of managed servers, the load balancer, and one of the management servers are installed in a data center, and the data center and the other management server are connected to a network. .. The instruction signal is a signal that excludes the application from the distribution target of the processing load, and is a predetermined time after the transmitter transmits the test signal for all the applications in the tests by both management servers. If the receiving unit does not receive the response signal even after the passage of time, or
The instruction signal is a signal that excludes the application from the distribution target of the processing load in the virtual server unit, and in the test by both the management servers, the transmission unit has the same with respect to at least one application in each of the virtual servers. When the receiving unit does not receive the response signal even after a predetermined time has passed since the test signal was transmitted, or
The instruction signal is a signal that excludes the application from the distribution target of the processing load for each physical server, and in the test by both the management servers, the transmission unit has the same with respect to at least one application in each of the physical servers. If the receiver does not receive the response signal within a predetermined time after transmitting the test signal,
The cloud system according to claim 10, wherein the instruction unit exceptionally does not output the instruction signal and sets the processing priority of some of the applications to be higher than the processing priority of the remaining applications. A management method for a cloud system including a management server, a plurality of managed servers, and a load balancer that distributes a processing load to the plurality of managed servers.
The management server
A transmission process of transmitting a test signal to a test communication terminal via each of the applications executed by the plurality of managed servers, and a transmission process.
A receiving step of receiving a response signal to the test signal transmitted from the test communication terminal via the application used at the time of transmitting the test signal.
When the receiving process does not receive the response signal even after a predetermined time has elapsed since the transmitting step transmitted the test signal, the application used at the time of transmitting the test signal may be used alone or. An instruction step of outputting an instruction signal to be excluded from the distribution target of the processing load to the load balancer together with the other application belonging to the same virtual server or together with the other application belonging to the same physical server.
A management method characterized by performing.

Images