JP2002251384A - Wide area cluster control system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a wide area cluster control system which has high reliability including disaster resistance and high availability by controlling a plurality of cluster systems dispersed in a wide area via a network and normally controlling the cluster systems. SOLUTION: The cluster system 1 which is actually operated is decided based on state information on the cluster system 1 transmitted from the system monitoring demons 11 (11a to 11n) of a plurality of the cluster systems 1 (1a to 1n). State information on a plurality of the cluster systems 1 (1a to 1n) is informed to the system control means (12a to 12n) of a plurality of the cluster systems (1a to 1n). The system control means (12a to 12n) of a plurality of the cluster systems 1 (1a to 1n) control a plurality of the cluster systems (1a to 1n) in accordance with the state information.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a cluster system which is remotely distributed, in which a client receives and stores the status, and notifies each cluster system of the status. It enables normal control and improves service availability.

[0002]

2. Description of the Related Art FIG.
5 is a network management method shown in FIG. 5, in which 102 and 104 are control point processors, 112 and 114 are disks, and 106 and 108 are customer processors. The control point processor and the customer processor are interconnected by a communication network 116. Each disk is connected to a control point processor. Further, the control point processors are classified into one main control processor and a spare control processor, and the disk of the spare control processor is synchronized with the disk of the main control processor.

Next, the operation will be described. The customer processor sends a request service task to the main control processor, and the main control processor returns a service response to the request. Here, the customer processor checks the service response, and if no service response is received, sends a request from the customer processor to one of the spare control processors to become the main control processor.

[0004]

As described above, in the conventional network management method, when the network 116 is vertically divided, the customer processor 106 regards the control point processor 102 as the main control processor, and the customer processor 108 Control point processor 104 is considered the main control processor. At this time, a plurality of main control processors exist in the network. If there is an application that operates exclusively in each control processor, there are a plurality of main control processors. For example, there is a problem that it is difficult to control the cluster system normally when a network failure occurs. Sometimes.

[0005] The present invention has been made to solve the above-mentioned problems, and a plurality of cluster systems distributed over a wide area are controlled via a network, so that the cluster systems can be controlled normally, thereby achieving disaster resistance. The purpose is to improve reliability and availability in consideration of the above.

[0006]

According to a first aspect of the present invention, there is provided the following element. (A) a plurality of cluster systems that are connected to a network and execute processing based on information transmitted from a client, and have the following elements; (a1) monitor the status of the cluster system, and System monitoring means for transmitting the status information and instructing the cluster system to stop when the status information cannot be transmitted; (a2) having a rule table for each of the plurality of cluster systems for controlling the plurality of cluster systems; A system control for controlling the plurality of cluster systems based on the state information of the plurality of cluster systems transmitted from the client and the rule table and stopping the cluster systems based on an instruction from the system monitoring means; Means; (b) connect to the above network Transmitting the status information transmitted by the plurality of cluster systems to the plurality of cluster systems, and storing a status information transmitted by the system monitoring means; and storing the status information transmitted by the system monitoring means. System information management means for deciding a cluster system to be actually operated when the status information which has been transmitted differs from the newly transmitted status information, and notifying the system control means of the plurality of cluster systems of the status information of the plurality of cluster systems .

A second invention comprises a plurality of clients each having the above-mentioned system information management means, wherein the above-mentioned system control means includes a plurality of cluster system status information transmitted from the plurality of clients, the above-mentioned rule table, Is used to control the plurality of cluster systems.

[0008] A third invention has the above-mentioned rule table, and the plurality of cluster systems constituted by the rule table and the status information of the plurality of cluster systems transmitted from the client or the plurality of clients. And a system control server for transmitting the operation information to the system control means. The system control means controls the plurality of cluster systems based on the operation information.

According to a fourth aspect of the present invention, there is provided a rule table having client information indicating whether or not the client transmitting status information is normal, the status information of the plurality of cluster systems transmitted from the client, and the rule table. And system control means for controlling the plurality of cluster systems based on the client information.

A fifth aspect of the present invention relates to a plurality of weighted clients for giving priority to the processing of the plurality of cluster systems, and the status information of the plurality of cluster systems transmitted from the weighted clients. System control means for controlling the plurality of cluster systems based on the rule table.

A sixth invention provides a rule table in which the rule tables for the plurality of cluster systems are integrated,
System control means for controlling the plurality of cluster systems based on the state information of the plurality of cluster systems transmitted from the client and the integrated rule table.

A seventh aspect of the present invention is a system information management means for transmitting the status information of a plurality of cluster systems and the system control information for causing the system control means to recognize the cluster system to execute the processing, System control means for controlling the plurality of cluster systems based on the status information and the system control information transmitted from the system information management means.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1 FIG. 1 is a configuration diagram of a wide-area cluster control method according to the first embodiment. In the figure, reference numeral 1 (1a to 1n) denotes a plurality of remotely distributed cluster systems, 2 denotes a client, and 3 denotes a wide area network connecting the plurality of cluster systems 1 and the clients 2. The business application is executed on any one of the cluster systems 1. Here, in each cluster system 1, the cluster system 1 in which the business application is being executed is a primary cluster system, and the rest are secondary cluster systems. Each cluster system 1
Is a system monitoring daemon 11 (11a to 11n) that collects status information indicating the system status of each cluster system 1 and sends it to the client 2, and a system control daemon 12 (12a to 12n) that controls the cluster system 1.
12n). In addition, each system control daemon 12
Have rule tables 13 (13a to 13n) used to control the cluster system 1, respectively. The client 2 transmits each cluster system 1 transmitted from the system monitoring daemon 11 of each cluster system 1.
It has a system information management daemon 21 for receiving status information of the system.

Next, the operation will be described. FIG. 2 is a flowchart illustrating the processing operation of the system monitoring daemon 11. Normally, each system monitoring daemon 11 monitors the status of the cluster system 1 having the system monitoring daemon 11 at regular intervals, and transmits the status information of the monitored cluster system 1 to the system information management daemon 21 of the client 2. (Step S1
~ S3). At this time, for example, the system monitoring daemon 1
When 1a confirms that a failure has occurred in the cluster system 1a, the system monitoring daemon 11a immediately notifies the system information management daemon 21 of the cluster system 1a.
Is notified that a failure has occurred (step S4).

When the failure of the cluster system 1a is notified to the system information management daemon 21 (step S5), the process waits until the failure of the cluster system 1a is restored (step S7), and the cluster system 1a Is recovered (step S8), the process returns to step S1, and the system monitoring daemon 11a monitors the status of the cluster system 1a again at regular intervals.

On the other hand, the system monitoring daemon 11a sends the system information management daemon 12 to the cluster system 1
If the notification that the failure has occurred in a cannot be made, the system monitoring daemon 11a instructs the system control daemon 12a to stop the cluster system 1a (step S6). After that, it waits until the failure of the cluster system 1a is recovered (step S
7) When the failure of the cluster system 1a recovers (step S8), the process returns to step S1, and the system monitoring daemon 11a monitors the state of the cluster system 1a again at regular intervals. Such an operation is performed by the cluster system 1
This is performed in a cluster system other than a.

FIG. 3 is a flowchart showing the processing operation of the system information management daemon 21. The system information management daemon 21 receives the status information indicating the system status of each cluster system 1 constituting each system monitoring daemon 11 transmitted from each system monitoring daemon 11 (step S11), and stores this status information internally. It accumulates (step S12). Next, the system information management daemon 21 compares the status information of each cluster system 1 accumulated so far with the status information newly transmitted from each system monitoring daemon 11, and determines whether or not these status information are different. A determination is made (step S13). When the status information does not differ, that is, when the status information is the same, the process returns to step S11, and the status information transmitted from the system monitoring daemon 11 is received again.

On the other hand, when it is determined in step S13 that these pieces of status information are different, the cluster system 1 that actually operates is determined, and the status information of each cluster system 1 is notified to each system control daemon 12 (step S13). S14).

Here, the actually operating cluster system 1 is determined based on a rule table as shown in FIG. A primary cluster system that is a cluster system 1 in which a business application is executed;
Consider a case where an abnormality occurs in the primary cluster system in an environment composed of two cluster systems of the remaining cluster system 1 and the secondary cluster system. In this case, if the secondary cluster system is normal or abnormal, it notifies the system control daemon 12 to stop the primary cluster system based on the rule table of the primary cluster system in FIG.

When an abnormality occurs in the primary cluster system and when the secondary cluster system is normal, the system control daemon 12 is notified to start the secondary cluster system based on the rule table of the primary cluster system in FIG. If an error occurs, the secondary cluster system is notified to stop.

The rule table of the primary cluster system in FIG. 4 is a rule table of the cluster system 1 on which the business application is executed, and the rule table of the secondary cluster system is other than the cluster system 1 on which the business application is executed. 3 is a rule table of the cluster system 1.

Next, when the system information management daemon 21 is able to send status information to the system control daemon 12
(Step S15), the system information management daemon 21 returns to step S11, and receives the state information transmitted from the system monitoring daemon 11 again.

On the other hand, if the system information management daemon 21 cannot send the status information to the system control daemon 12 in step S15, it is determined that the failed cluster system 1 is down (step S16). Then, the fact that the cluster system 1 is down is stored as state information of the new cluster system 1.

FIG. 5 is a flowchart showing the processing operation of the system control daemon 12. Each system control daemon 12 determines whether or not status information of each cluster system 1 has been received from the system information management daemon 21 of the client 2 (step S21), and if status information has been received from the system information management daemon 21, The respective cluster systems 1 are controlled in accordance with the received status information and the rule table 13 (step S25), and thereafter, the process returns to step S21 to determine whether or not status information has been received from the system information management daemon 21 of the client 2 again. I do.

On the other hand, if the status information cannot be received from the system information management daemon 21 in step S21, a predetermined interval is waited (step S22), it is determined whether or not a timeout has occurred (step S23), and the timeout has occurred. When it is determined that the cluster system 1 having the status information that could not be received from the system information management daemon 21 determines that the service cannot be performed, the cluster system 1 of the system control daemon 12 that could not receive the status information is stopped (step S24). Then, returning to step S21,
It is determined again whether the status information notified from the system information management daemons 21 of all the clients 2 has been received.

If it is determined in step S23 that the timeout has not occurred, the process returns to step S21 to determine whether the status information notified from the system information management daemon 21 of the client 2 has been received. FIG.
Shows operations of the system control daemon 12, the system monitoring daemon 11, and the system information management daemon 21 when a failure occurs in the cluster system 1.

As described above, according to the present embodiment, when a failure occurs in the cluster system 1 such as a disaster or when a certain cluster system 1 and the wide area network 3 are separated, the system information management of the client 2 is performed. The cluster system 1 that cannot communicate with the daemon 21 stops,
Only one of the cluster systems 1 that can communicate with the system information management daemon 21 is started. As a result, when a failure occurs in the cluster system 1 such as a disaster or a network
Can be controlled normally. As a result, a wide-area cluster control method with high reliability and high availability considering disaster resistance can be realized.

Embodiment 2 In the first embodiment, the case where the state information of the cluster system is accumulated using a single client has been described. In the present embodiment, a case where the reliability is improved by using a plurality of clients will be described.

FIG. 7 is a configuration diagram of the wide area cluster control system according to the second embodiment. In the figure, 1 (1a to 1n) is a plurality of cluster systems distributed remotely, and 2 (2a to 2
m) is a plurality of clients, 3 is a plurality of cluster systems 1 (1a to 1n) and a plurality of clients 2 (2a to 2n).
m) is a wide area network. Each cluster system 1 collects status information indicating the system status of each cluster system 1 and transmits the collected status information to all the clients 2, and a system control daemon 12 that controls the cluster system 1. (12a to 12n). Also, each system control daemon 12 has a rule table similar to the rule table shown in FIG. 4, that is, a rule table 13 (13a to 13a) used for controlling the cluster system 1.
13n). Each client 2 has a system information management daemon 21 (21a to 21m) that receives status information of each cluster system 1 transmitted from the system monitoring daemon 11 of each cluster system 1. The business application is executed on any one of the cluster systems 1.

Next, the operation will be described. FIG. 8 is a flowchart showing the processing operation of the system monitoring daemon 11. The flowchart of FIG. 8 is the same as the flowchart of FIG. 2 except that each system monitoring daemon 11 transmits the status information of each cluster system 1 to all the clients 2 in step S34. Also, the system information management daemon 21 of each client 2
Is the same as that of the system information management daemon 21 described in the first embodiment.

FIG. 9 is a flowchart showing the processing operation of the system control daemon 12. Each system control daemon 12 determines whether or not status information of the cluster system 1 has been received from the system information management daemons 21 of all the clients 2 (step S41). In step S47, the cluster system 1 is controlled in accordance with the received state information and the rule table 13.
Returning to step S41, it is determined again whether or not status information has been received from the system information management daemons 21 of all the clients 2.

On the other hand, when the status information cannot be received from the system information management daemon 21, a predetermined interval is waited (step S42), it is determined whether or not a timeout has occurred (step S43), and it is determined that the timeout has occurred. Sometimes, the system information management daemon 2 that could not receive
It is determined that the first client 2 has stopped, and is excluded from the reception target (step S44). Thereafter, the number of clients to be received is checked (step S45).
Is determined. If there are no more clients to receive, the cluster system determines that the service is not possible and stops the cluster system (step S).
46). Thereafter, the process returns to step S41, and it is determined again whether or not the status information notified from the system information management daemons 21 of all the clients 2 has been received.

If it is determined in step S43 that a timeout has not occurred, the process returns to step S41, and it is determined again whether information on the system information management daemon 21 of all clients 2 has been received. As described above, the system control daemon 12 records the number of clients 2 connected to the wide area network 3 in advance, summarizes the received information from the plurality of clients 2 by majority decision, and configures the cluster system according to the rule table. Control.

As described above, according to the present embodiment, even if a failure occurs in the cluster system 1 such as a disaster, or if a certain cluster system 1 and the wide area network 3 are disconnected, The cluster system 1 can be controlled normally, and the reliability of the state information of each cluster system 1 can be improved. Further, a wide-area cluster control method with high reliability and high availability in consideration of disaster resistance can be realized.

Embodiment 3 FIG. In the second embodiment, the storage destination of the status information from the system information management daemon is the system control daemon of each cluster system. However, in the present embodiment, another computer is installed, and the system information management daemon is installed in this computer. A description will be given of a case where the state information is stored.

FIG. 10 is a configuration diagram of the wide area cluster control system according to the third embodiment. In the figure, 1 (1a to 1n)
Is a plurality of cluster systems distributed remotely, and 2 (2a to
2m) is a plurality of clients, 3 is a plurality of cluster systems 1 (1a to 1n) and a plurality of clients 2 (2a to 2n).
2m). Also, each cluster system 1 collects status information indicating the system status of each cluster system 1, and collects all the clients 2
System monitoring daemon 11 (11a to 11a)
n) and a system control daemon 12 (12a to 12n) for controlling the cluster system 1.

The system control server 4 is connected to the wide area network 3, and each client 2 receives the status information transmitted from the system monitoring daemon 11 of each cluster system 1 and sends the status information to the system control server 4. System information management daemon 21 (2
1a to 21m). Further, the system control server 4
Has a rule table 13 similar to the rule table 13 of each system control daemon 12 shown in FIG.

Next, the operation will be described. The processing operation of each system monitoring daemon 11 is the same as the flowchart of FIG. In addition, the processing operation of each system information management daemon 21 is described as “notifying the system control daemon 12 of the status information of the cluster system 1” in step S14 of FIG. The operation flow is the same as that described with reference to FIG.

FIG. 11 is a flowchart showing the processing operation of the system control server 4. Each system control server 4 determines whether the status information of the cluster system 1 has been received from the system information management daemons 21 of all the clients 2 (step S51), and when the status information has been received from each system information management daemon 21. Accumulates the received status information from each information management daemon 21,
The operation information of the cluster system 1 composed of the accumulated state information and the rule table 13 of the cluster system 1 is transmitted to the system control daemon 12 of the cluster system 1 (step S55). Thereafter, the process returns to step S51, and it is determined again whether or not status information has been received from the system information management daemons 21 of all the clients 2. When the system control server 4 transmits the operation information of the cluster system 1 to the system control daemon 12 in step S55, the system control daemon 12 controls the cluster system 1 based on the operation information.

On the other hand, if the status information is not received from the system information management daemon 21 in step S51, a predetermined interval is waited (step S52), it is determined whether or not a timeout has occurred (step S53), and it is determined that a timeout has occurred. In this case, the client 2 having the information of the system information management daemon 21 that cannot be received is determined to be stopped, and is excluded from the reception target (step S5).
4). Thereafter, the process returns to step S51, and it is determined again whether or not status information has been received from the system information management daemons 21 of all the clients 2.

That is, in the present embodiment, the system control server 4 receives the information from the system information management daemon 21 and converts the received information and the operation information of the cluster system 1 obtained from the rule table 13 into each cluster. This is sent to the system control daemon 12 of the system 1. The system control daemon 12 controls the cluster system according to the operation information obtained from the system control server 4.

As described above, according to the present embodiment, even if a failure occurs in the cluster system 1 which is a disaster or the like, or if a certain cluster system 1 is separated from the wide area network 3, each client 2 Cluster system 1
It is not necessary to perform communication with the network, and the availability of the network can be improved. In addition, a wide area cluster control method with high reliability and high availability in consideration of disaster resistance can be realized.

Embodiment 4 FIG. In the second embodiment, the operations of the primary cluster system and the secondary cluster system are determined by the majority decision of the notification from the client and the rule table. However, in the present embodiment, the cluster system 1 is determined based on the weighted rule table. Will be described.

FIG. 12 is a diagram showing an example of a weighted rule table. FIG. 12 is a table showing that the operation of the cluster system is switched when 60% or more of the total number of clients is considered normal. The configuration of the wide area cluster control system of the present embodiment and the processing operations of the components are the same as those of the second embodiment. The present embodiment is different from the second embodiment in that the cluster system 1 that actually operates is determined based on the weighted rule table in FIG.

Thus, the operation of the cluster system can be adjusted according to the network environment of the system to which the system is applied. In other words, it is possible to realize a cluster system that operates as long as there is a client that regards the cluster system as normal even if there is at least one client, and a cluster system that stops if there is at least one client that regards the cluster system as abnormal. .

As described above, according to the present embodiment, it is possible to reduce unexpected system operation due to erroneous recognition of the system information management daemon 21 by using the weighted rule table, and Can be improved in reliability.

Embodiment 5 In the second embodiment, the status information transmitted from the system information management daemon of each client to the system control daemon of each cluster system is equal for each client. A case will be described in which weighting is performed, that is, weighting is performed so as to give priority to processing of a plurality of cluster systems.

The configuration of the wide area cluster control system of the present embodiment and the processing operation of the components are the same as those of the second embodiment. In the present embodiment, when the system control daemon 12 determines the operating state of the cluster system 1 based on the rule table 13, the result of a specific client 2 is preferentially reflected. That is, the system control daemon 21 controls each cluster system 1 based on the status information of each cluster system 1 transmitted from a specific client 2 and the rule table 13.

According to the present embodiment, it is possible to lower the priority of a low-reliability client or a client connected to a low-reliability network path. Can be minimized.

Embodiment 6 FIG. In the first to fifth embodiments, it is necessary to prepare the same number of rule tables as the number of cluster systems. In the present embodiment, a case will be described in which a plurality of rule tables are combined into one. In this case, one cluster system becomes a primary cluster system, and the remaining cluster systems become secondary cluster systems.

FIG. 13 is a diagram showing an example of a rule table obtained by integrating a plurality of rule tables into one. The configuration of the wide area cluster control system of this embodiment and the processing operations of the components are the same as those of the first to fifth embodiments, but there are differences in the processing operations of the system control server 4 as described below.

The processing operation of the system control daemon 12 is as follows. Each of the system control daemons 12 sends the cluster system 1a, which is the primary cluster system, from the system information management daemon 21 of all the clients 2.
Is stopped, the cluster system 1a, which is the primary cluster system, is demoted to the secondary cluster system, and the next primary cluster system among the other cluster systems that are operating normally is replaced with the next primary cluster system. Select a cluster system. Then, according to the rule table 13, the cluster system 1
Control.

When the system control server 4 is used, the same processing as that of the system control daemon 12 is performed as described above. However, when the system control server 4 is used, it operates normally. When the next primary cluster system is selected from other cluster systems, which are secondary cluster systems, the operation information of the cluster system 1 composed of this information and the rule table 13 of the cluster system 1 is transmitted to the system control of the cluster system 1. The system control daemon 12 controls the cluster system in accordance with the operation information.

As described above, according to the present embodiment, there is no need to manage a plurality of rule tables, it is possible to improve the availability of the system, and to realize a highly available wide area cluster control system. Can be.

Embodiment 7 FIG. In the first to sixth embodiments, the client transmits the status information of each cluster system to the system control daemons of all the cluster systems. However, in the present embodiment, when a high-speed network is used, A case where status information and system control information are transmitted only to the cluster system that is the primary cluster system will be described. In this case, the cluster system of the system control daemon that has received the status information and the system control information automatically becomes the primary cluster system.

The configuration of the wide area cluster control system of this embodiment and the processing operation of this component are the same as those of the first to sixth embodiments, but the following operations are performed in the first to sixth embodiments.
It is a different point.

The operation will be described. The system information management daemon 21 of the client 2 periodically starts the system control daemon 12 of the specific cluster system 1,
For example, the system control daemon 1 of the cluster system 1a
2a, the status information of each cluster system 1 and the system control information for recognizing the cluster system as the primary cluster system are transmitted. The system control daemon 12a of the cluster system 1a, which is the primary cluster system, confirms that it is the primary cluster by receiving status information and system control information transmitted at regular intervals. If control information cannot be received, demote from the primary cluster to the secondary cluster. When the status information and the system control information cannot be transmitted to the cluster system 1a as the primary cluster system, the system information management daemon 21 determines the status information and the system control to determine a new primary cluster based on the cluster control algorithm described above. The information is transmitted to the system control daemon 12 of another cluster system 1.

As described above, according to the present embodiment, the system information management daemon 21 of each client 2 does not need to send status information to the system control daemons 12 of all the cluster systems 1, and the system Can be improved, and a wide-availability cluster control method with high availability can be realized.

[0059]

Since the present invention is configured as described above, it has the following effects.

In the first invention, the cluster system to be actually operated is determined based on the status information of the cluster systems transmitted from the system monitoring means of the plurality of cluster systems, and the plurality of cluster systems are determined by the system control means of the plurality of cluster systems. The system control means of a plurality of cluster systems controls the plurality of cluster systems in accordance with the status information, thereby causing a failure in the cluster system such as a disaster or a failure in the network. When a problem occurs, each cluster system can be controlled normally, and wide area cluster control with high reliability and high availability in consideration of disaster resistance can be realized.

In the second invention, by providing a plurality of clients each having system information management means, it is possible to perform cluster control with further improved reliability, reliability and availability in consideration of disaster resistance. High-wide-area cluster control can be realized.

According to a third aspect of the present invention, there is provided a system control server for transmitting operation information of a plurality of cluster systems composed of status information of a plurality of cluster systems transmitted from a client or a plurality of clients to system control means. This eliminates the need for each client to communicate with each cluster system, thereby improving network availability.

According to the fourth aspect of the invention, the operation of the cluster system can be adjusted according to the network environment of the system to be applied by having the client information indicating whether or not the client transmitting the state information is normal in the rule table. This makes it possible to realize high-availability wide-area cluster control.

According to the fifth aspect of the present invention, by giving a weight to a plurality of clients to execute the processes of a plurality of cluster systems preferentially, erroneous determination of a system control daemon due to a failure on the client side can be minimized. .

In the sixth aspect of the present invention, by integrating rule tables for a plurality of cluster systems into the above-mentioned rule table, there is no need to manage a plurality of rule tables, and high-availability wide-area cluster control can be realized. it can.

In the seventh invention, the system information management means transmits the state information of the plurality of cluster systems and the system control information for causing the system control means to recognize that the system is a cluster system for executing the processing.
The client system information management daemon does not need to send status information to the system control daemons of all cluster systems, and can realize highly available wide area cluster control.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a wide area cluster control method according to a first embodiment.

FIG. 2 is a flowchart showing a processing operation of a system monitoring daemon 11 according to the first embodiment.

FIG. 3 is a flowchart showing a processing operation of a system information management daemon 21 according to the first embodiment.

FIG. 4 is a diagram showing an example of a rule table according to the first embodiment.

FIG. 5 is a flowchart showing a processing operation of a system control daemon 12 according to the first embodiment.

FIG. 6 is a diagram showing an operation of each daemon when a failure occurs in the cluster system 1 in the first embodiment.

FIG. 7 is a configuration diagram of a wide-area cluster control method according to the second embodiment.

FIG. 8 is a flowchart showing a processing operation of a system monitoring daemon 11 according to the second embodiment.

FIG. 9 is a flowchart showing a processing operation of a system control daemon 12 according to the second embodiment.

FIG. 10 is a configuration diagram of a wide-area cluster control method according to the third embodiment.

FIG. 11 is a flowchart showing a processing operation of a system control server 4 according to the third embodiment.

FIG. 12 is a diagram showing an example of a weighted rule table according to the fourth embodiment.

FIG. 13 is a diagram showing an example of a rule table obtained by integrating a plurality of rule tables into one according to the sixth embodiment.

FIG. 14 is a configuration diagram of a conventional network management method.

[Description of Signs] 1 (1a to 1n) cluster system, 2 clients, 3 wide area network, 11 (11a to 11n)
System monitoring daemon, 12 (12a to 12n) System control daemon, 13 (13a to 13n) Rule table.

Claims (7)

[Claims] 1. A wide-area cluster control method having the following elements. (A) a plurality of cluster systems that are connected to a network and execute processing based on information transmitted from a client, and have the following elements; (a1) monitor the status of the cluster system, and A system monitoring means for transmitting the status information and instructing to stop the cluster system when the status information cannot be transmitted; (a2) having a rule table for each of the plurality of cluster systems for controlling the plurality of cluster systems; System control for controlling the plurality of cluster systems based on the status information of the plurality of cluster systems transmitted from the client and the rule table, and for stopping the cluster systems based on an instruction from the system monitoring means; Means; (b) connect to the above network Transmitting the status information transmitted by the plurality of cluster systems to the plurality of cluster systems, and storing the status information transmitted by the system monitoring means; and (b1) storing the status information transmitted by the system monitoring means. System information management means for deciding a cluster system to be actually operated when status information that has been transmitted differs from newly transmitted status information, and notifying the system control means of the plurality of cluster systems of the status information of the plurality of cluster systems . 2. A system comprising: a plurality of clients each having the system information management unit; wherein the system control unit is configured to execute the plurality of cluster systems based on the status information of the plurality of cluster systems transmitted from the plurality of clients and the rule table. 2. The wide area cluster control method according to claim 1, wherein a plurality of cluster systems are controlled. 3. The system according to claim 2, further comprising: a rule table, wherein operation information of the plurality of cluster systems, comprising the rule table and the status information of the plurality of cluster systems transmitted from the client or the plurality of clients, is stored. 3. The wide area cluster according to claim 1, further comprising: a system control server for transmitting to the system control means, wherein the system control means controls the plurality of cluster systems based on the operation information. control method. 4. The rule table has client information indicating whether the client transmitting status information is normal or not, and the system control means transmits status information of the plurality of cluster systems transmitted from the client. And controlling the plurality of cluster systems based on the client information in the rule table.
The described wide area cluster control method. 5. The plurality of clients are weighted so as to give priority to the processing of the plurality of cluster systems, and the system control means transmits a status of the plurality of cluster systems transmitted from the weighted client. 5. The wide area cluster control method according to claim 4, wherein said plurality of cluster systems are controlled based on information and said rule table. 6. The rule table is integrated with a rule table for each of the plurality of cluster systems, and the system control means includes a state information of the plurality of cluster systems transmitted from a client, and the integrated rule table. 6. The wide area cluster control method according to claim 5, wherein the plurality of cluster systems are controlled based on the following. 7. The system information management means, wherein the system control means stores status information of a plurality of cluster systems,
Transmitting system control information for recognizing that the plurality of clusters is a cluster system that executes processing, wherein the system control unit transmits the plurality of clusters based on the state information and system control information transmitted from the system information management unit; 7. The wide area cluster control system according to claim 1, wherein the system is controlled.