JP2003256399A - Control method for switching in hot standby system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a control method for switching in a hot standby system of a loosely-coupled multiprocessor comprising more than three computers, which levels out loads of each computer. <P>SOLUTION: A first computer 100 switches a transactional application operated by the first computer 100 to a second computer 200 as an operation for the second computer 200 when the first computer 100 fails. A means 206 to notify switching of a batch application notifies the switching of the batch application to a third computer 300, if a means 205 to judge the switching of the batch applications judges that an application load of the second computer 200 becomes more than the application load of the third computer 300. A first means 302 for switching the batch applications processes a switching process in order to hand over the batch applications to a first means 301 for alternative processing. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a hot standby system switching control system, and more particularly to a hot standby system switching control system in a loosely coupled multiprocessor system in which three or more computers are channel-connected.

[0002]

2. Description of the Related Art An example of a conventional hot standby system is described in Japanese Patent Laid-Open No. 5-165664.
As shown in FIG. 12, this conventional hot standby system includes an active computer 1 and an alternate computer 2,
The operating system computer 1 monitors the business processing means 3 for executing transaction processing and whether the business processing means 3 is operating normally, and at the time of normal operation, a normal message indicating a normal operation at regular monitoring time intervals. When a failure occurs in which the transaction processing cannot be continued in the normal message sending means 5 and the business processing means 3, the failure message sending means 8 detects this failure and sends a failure message indicating this failure. The alternative computer 2 receives the normal business message from the alternative business processing means 4 that takes over the transaction processing and the normal business message sending means 5, and the business processing means 3 is operating normally by this reception. The monitoring message receiving means 6 for recognizing that the failure message is received from the failure message transmitting means 8 and outputting the failure message, and the monitoring message receiving means 6 Receives a harm notification, analyzes it, determines the occurrence of a failure in the business processing means 3, receives a failure notification from the monitoring message analysis means 9 for notifying the failure, and receives a failure notification from the monitoring message analysis means 9, thereby performing alternative business processing. Business switching control means 7 for switching transaction processing to the means 4
It was equipped with and.

The conventional hot standby system having such a configuration operates as follows.

In the operational computer 1, the business processing means 3 is
Process transactional business and execute this transactional business. The normal message transmitting means 5 monitors whether or not the business processing means 3 is operating normally in the operational computer 1, and when it is operating normally, substitutes a normal message indicating normal operation via the communication line. It is transmitted to the system computer 2 at regular monitoring time intervals. Fault message transmitting means 8
Monitors the occurrence of a failure in the business processing means 3, and when a failure is detected, transmits a failure message indicating the failure to the alternative computer 2.

In the alternative computer 2, the monitoring message receiving means 6
When the normal message is received, the business processing means 3 recognizes that it is operating normally. If the normal message is not received, it is considered that a failure has occurred in the operating system computer 1, and the failure is switched to the business switching control means. Notify 7. Further, the monitoring message receiving means 6 outputs the failure message to the monitoring message analyzing means 9 when receiving the failure message indicating the failure from the failure message sending means 8. The monitoring telegram analysis means 9 determines the fault of the operating system computer 1 by analyzing the input fault telegram, and gives an instruction to the work switching control means 7. Business switching control means 7
When the monitoring message receiving means 6 or the monitoring message analyzing means 9 notifies the failure of the active computer 1, the transaction work is switched to the alternative work processing means 4. The alternative business processing means 4 continues the business processing of the transaction when a failure occurs in the operating system computer 1, and executes the processing of the succession.

The hot standby system described in Japanese Patent Laid-Open No. 5-165664 is composed of two computers, an active system computer and an alternative system computer. A hot standby system including the above host computer is disclosed.

In the standby system allocation host determination method in the hot standby system disclosed in Japanese Patent Laid-Open No. 07-073061, the presence / absence of the active system and standby system of the logical system and the logical system of each active system is monitored for each host computer. Logical system status monitoring device, inter-host communication device that enables mutual communication between host computers, host load number storage device for storing the host load number given at the time of system installation in advance, and inter-host communication A host load calculation device that calculates the load of each host computer that configures the hot standby system based on the information obtained through the device and the information that exists in the host load number storage device and stores the load in the host load status storage device,
It is composed of an operation monitoring terminal for taking out load information of each host from the host load status storage device and allowing a system operator to instruct and inquire the operation of the system.

In the conventional hot standby system configured as described above, the system operator can be informed which host is most suitable to start up the standby system of the logical system, so that the load balance of the entire system can be determined. The logical system can be arranged appropriately and dynamically.

[0009]

However, JP-A-5-1 is used.
In the prior art disclosed in Japanese Patent No. 65664, when a transaction operation is switched, the operation load of the alternative computer thereafter becomes high in the alternative computer, so that the transaction operation and the batch operation in the subsequent alternative computer are performed. However, there was a problem that there was a possibility of delay and performance degradation in the operation of.

Further, in the prior art disclosed in Japanese Patent Laid-Open No. 07-073061, the system operator requests the display of the host load coefficient at the time of starting the logical system, and the host computer for starting the logical system is determined based on the display. However, there is a problem that a wrong logical system may be arranged due to human intervention.

An object of the present invention is to provide a hot standby system consisting of a loosely coupled multiprocessor system in which three or more computers are connected in a channel, and in the first computer when a failure occurs in the first computer. If the transaction work being operated is switched to the second computer and the load on the second computer becomes high,
Another object of the present invention is to provide a hot standby system switching control method in which a batch job that has been operated on this computer is switched to the third computer.

[0012]

The hot standby system switching control method of the present invention is a loosely coupled multiprocessor system in which three computers are connected in channels, and transaction operations are performed during normal business operations. When the operating system to be executed exists on the first computer and the failure that occurred on the first computer makes it impossible to continue the transaction work on the operating system, restore the failed transaction work. In a hot standby system with a configuration in which a second computer having an alternative system that takes over and executes the subsequent transaction operations and a third computer are connected, the first computer operates during normal business operations. A failure occurred in the transaction business processing means that executes transaction business as a system and the first computer In the case where it is determined by the failure event determination means that the transaction work in the active system cannot be continued, the failure event determination means for determining whether or not the transaction work in the active system can be continued A failure notification means for notifying a transaction operation switching means existing in another computer of the failure and instructing switching of the transaction operation, and the second computer receives the operation instruction for switching the transaction operation from the failure notification means, and the operating system Transaction transaction switching means that restores the transaction transaction that was being executed in a state in which execution can be continued, transaction transaction alternative processing means that continues execution by succeeding and executing subsequent transaction transactions after recovery of transaction transactions, Batch business process that executes batch business during operation And a first work load information monitoring means for monitoring the work load of the second computer at constant monitoring time intervals and recording the work load information of the second computer in the work load information table, and the first computer. When a failure occurs, the execution of the transaction work is switched and the transaction work is resumed by the transaction work alternative processing means, the work load information table is referred to and the work load information of the second computer and the work of the third computer are referred to. A batch job switching determination means for comparing the load information, and determining whether to switch the batch job to the third computer when it is detected that the business load of the second computer has increased; And a batch operation switching notifying means for instructing the third computer to switch the batch operation when it is determined to switch the batch operation to the third computer.
A second work load information monitoring means for monitoring the work load of the third computer at a constant monitoring time interval and recording the work load information of the third computer in the work load information table; and the batch work. Batch job switching processing means for switching a batch job instructed to be switched by the switching notifying means from the second computer, and batch job alternative processing means for continuing execution of the batch job switched by the batch job switching processor And is provided.

Further, in the hot standby system switching control method of the present invention, the work load information of the second computer and the work load information of the third computer are recorded in the work load information table in correspondence with each time. It is characterized by

Further, the hot standby system switching control method of the present invention is a loosely coupled multiprocessor system in which four or more computers are connected in channels, and during normal business operation, transaction business is executed. The first computer where the system exists, and the transaction after the transaction that caused the failure is restored when the transaction in the active system cannot be continued due to the failure that occurred in the first computer In a hot standby system with a configuration in which a second computer that has an alternative system for taking over and executing business operations and two or more third computers are connected, the first computer is Transaction business processing means for executing transaction business as an active system, and first
Failure event determining means for determining whether or not it is possible to continue the transaction work in the active system when a failure occurs in the computer, and it is impossible to continue the transaction work in the active system by the failure event determining means. And a failure notification means for notifying the transaction work switching means existing in the second computer of the failure and instructing switching of the transaction work when the determination is made. Transaction business switching means that receives a switching instruction and restores the transaction business that was being executed in the active system to a state in which it can continue execution, and transaction business that continues operation by succeeding and executing subsequent transaction business after recovery of the transaction business Alternate processing means and batch operations during normal operation A batch work processing means to be executed; a first work load information monitoring means for monitoring the work load of the second computer at a constant monitoring time interval and recording the work load information of the second computer in the work load information table; When a failure occurs in the first computer, the execution of the transaction work is switched, and the transaction work alternative processing means resumes the transaction work, the work load information table is referred to refer to the work load information of the second computer and The third computer is compared with the third computer's workload information, and when it is detected that the second computer's workload is high, the third computer identifies the one with the lowest workload and assigns it to the third computer. Batch job switching determination means for determining whether or not to switch the batch job, and batch job when it is determined to switch the batch job to the third computer. And a batch work switching notification means for instructing the third computer to switch the work load of the third computer, and the third computer monitors the work load of the third computer at constant monitoring time intervals, and the work load information of the third computer. In the work load information table, a batch work switching processing unit for switching the batch work instructed to be switched by the batch work switching notifying unit from the second computer, and the batch work. It is characterized in that it comprises a batch job alternative processing means for continuing the execution after the handover of the batch job switched by the switching processing means.

Furthermore, in the hot standby system switching control method of the present invention, the work load information table includes the work load information of the second computer and the work load information of two or more third computers at each time. It is characterized in that it is recorded corresponding to.

On the other hand, the hot standby system switching control method of the present invention is a loosely coupled multiprocessor system in which three computers are connected in a channel.
During normal business operation, if the first computer on which the active system that executes the transaction business exists and the failure that occurred on the first computer makes it impossible to continue the transaction business on the active system A second computer having an alternative system for recovering the failed transaction operation and taking over the operation of the subsequent transaction operation;
In a hot standby system having a configuration in which a third computer is connected, if a failure occurs in the first computer and it becomes impossible to continue the transaction work on the first computer, the second computer In this case, the transaction business is restored and the operation of the transaction business is taken over from the first computer, and when the business load of the second computer becomes high, the batch business executed by the second computer Is switched to a third computer and executed.

The hot standby system switching control method of the present invention is a loosely-coupled multiprocessor system in which four or more computers are channel-connected, and an operation for executing transaction work during normal business operation. The first computer in which the system exists, and the transaction after the transaction that caused the failure is restored when it becomes impossible to continue the transaction business in the active system due to the failure that occurred in the first computer In a hot standby system having a configuration in which a second computer having an alternative system for taking over and executing business operations and two or more third computers are connected, a failure occurs in the first computer, If it becomes impossible to continue the transaction work on this computer, the relevant transaction will be executed on the second computer. The business is restored and the operation of the transaction business is taken over from the first computer. When the business load of the second computer becomes high, the third computer identifies the one with the low business load, It is characterized in that the batch job executed by the second computer is switched to the third computer and executed.

On the other hand, the program of the present invention is a transaction business processing means for executing a transaction business as an active system of a computer during a normal business operation; Failure in the failure event determination means for determining whether or not continuation is possible, and the transaction operation switching means existing in another computer when the failure event determination means determines that the transaction operation in the active system cannot be continued Is notified and is operated as a failure notification means for instructing switching of transaction work.

Further, the program of the present invention is a transaction for recovering a computer to a state in which it is possible to continue execution of the transaction work executed in the active system in response to an instruction to switch the transaction work from a failure notification means of another computer. Business switching means, transaction business alternative processing means that takes over and executes subsequent transaction business after recovery of transaction business, batch business processing means that executes batch business during normal operation, and constant business load of the computer Monitoring means at a monitoring time interval of 1) to record the workload information of the computer in the workload information table, and the execution of the transaction job is switched when a failure occurs in the first computer and the transaction job replacement is performed. Tiger by processing means When the transaction work is restarted, the workload information of the computer of the computer is compared with the workload information of another computer by referring to the workload information table, and it is detected that the workload of the computer becomes high. Batch job switching determination means for determining whether to switch the batch job to another computer when it is performed, and switching the batch job to another computer when it is determined to switch the batch job to the other computer. It is characterized in that it is operated as a batch work switching notification means for instructing a computer.

Furthermore, the program of the present invention is a transaction for recovering a computer to a state in which it is possible to continue execution of a transaction job executed in an active system in response to an instruction to switch transaction jobs from a failure notification means of another computer. Business switching means, transaction business alternative processing means that takes over and executes subsequent transaction business after recovery of transaction business, batch business processing means that executes batch business during normal operation, and constant business load of the computer Monitoring means at a monitoring time interval and records the work load information of the computer in the work load information table. The first work load information monitoring means switches the execution of the transaction work due to a failure in the first computer Tran by Means When the operation work is restarted, the work load information table is referred to and the work load information of the computer of the computer is compared with the work load information of another computer to detect that the work load of the computer has increased. When this occurs, another computer with a low work load is identified, batch work switching determination means for determining whether to switch the batch work to the other computer, and batch work is switched to the other computer When it is determined that the batch job switching is instructed, the batch job switching notification means for instructing the other computer is operated.

Furthermore, the program of the present invention is a second work load information monitor for monitoring the work load of the computer at a constant monitoring time interval and recording the work load information of the computer in the work load information table. Means, a batch job switching processing means for switching a batch job instructed to be switched from another computer from the other computer, and a batch job alternative for continuing execution after the batch job switched by the batch job switching processing means It is characterized in that it is operated as a processing means.

Further, the program of the present invention is a transaction business processing means for executing transaction business as an active system in a computer during normal business operation, and a transaction business processing means for operating the transaction business in the active system when a failure occurs in the computer. Failure event determining means for determining whether or not continuation is possible, and failure by the transaction operation switching means existing in another computer when the failure event determination means determines that transaction operation in the active system cannot be continued Failure notification means for notifying and instructing the switching of transaction business, and transaction business switching means for receiving the transaction business switching instruction from the failure notification means and restoring the transaction business being executed in the active system to a state where execution can be continued ， Transaction industry After the recovery, the transaction transaction alternative processing means that inherits and executes the subsequent transaction transactions and continues operation, the batch transaction processing method that executes batch transactions during normal operation, and the workload of other computers at fixed monitoring intervals. A first work load information monitoring unit that monitors and records work load information of the other computer in a work load information table; execution of the transaction work is switched due to a failure in the computer, and transaction work is executed by the transaction work alternative processing unit. When the restart is restarted, the workload information table of the other computer is compared with the workload information of another computer by referring to the workload information table, and it is detected that the workload of the other computer has increased. When the job is executed, the batch job is switched to another computer. Batch job switching determination means for deciding whether or not to carry out a batch job switch notification means for instructing the other computer to switch the batch job when it is judged to switch the batch job to the other computer, the computer Second work load information monitoring means for monitoring the work load of the computer at a constant monitoring time interval and recording the work load information of the computer in the work load information table, and the batch work instructed to be switched by another computer. It is characterized in that it is operated as a batch job switching processing means for switching from the computer and a batch job alternative processing means for continuing the execution after the handover of the batch job switched by the batch job switching processing means.

Further, the program of the present invention is a transaction business processing means for executing a transaction business as an active system during a normal business operation of a computer, and a transaction business processing means for executing the transaction business in the active system when a failure occurs in the computer. Failure event determining means for determining whether or not continuation is possible, and failure by the transaction operation switching means existing in another computer when the failure event determination means determines that transaction operation in the active system cannot be continued Failure notification means for notifying and instructing the switching of transaction business, and transaction business switching means for receiving the transaction business switching instruction from the failure notification means and restoring the transaction business being executed in the active system to a state where execution can be continued ，
Subsequent transaction service alternative processing means that continues the operation by succeeding and executing the subsequent transaction business after recovery of the transaction business, batch business processing means that executes the batch business during normal operation, and a certain monitoring time for the workload of other computers Monitoring at intervals and recording work load information of the other computer in the work load information table 1st
The work load information monitoring means of the other computer refers to the work load information table when the execution of the transaction work is switched due to a failure in the computer and the transaction work is resumed by the transaction work alternative processing means. When the load information and the work load information of another computer are compared and it is detected that the work load of the other computer has increased, another computer with a low work load is identified and Batch job switching determination means for deciding whether or not to switch batch jobs, and batch job switch notification means for instructing another computer to switch batch jobs when it is determined to switch batch jobs to the other computer , Monitoring the work load of the computer for a certain period Second work load information monitoring means for remotely monitoring the work load information of the computer and recording the work load information of the computer in the work load information table, batch work switching processing for switching the batch work instructed to be switched by another computer from the other computer And a batch job switching processing means for operating the batch job switched by the batch job switching processing means as a batch job substitute processing means for continuing execution after the handover.

In FIG. 1, the hot standby switching system of the present invention is a loosely coupled multiprocessor system in which three or more computers are connected in a channel and executes transaction business during normal business operation. When the first computer 100 in which the active system exists and the transactional business become unable to continue the business in the active system due to a failure that occurred in the first computer 100, Restore,
In addition to a second computer 200 having an alternative system that takes over the operation of the subsequent transaction work and a computer having an active system and a computer having an alternative system, one or more third computers 300 are connected. In the first computer 100, which is a hot standby system having a configured
When a failure occurs in the transaction business processing means 101 that executes transaction business as an active system and the first computer 100 during normal business operation, it is determined whether or not the transaction business can be continued in the active system. Failure event determination means 102 and failure event determination means 102
However, when it is determined that the transaction business in the active system cannot be continued, the failure notification means 1 for notifying the transaction business switching means 202 existing in the second computer 200 of the failure and instructing the switching of the transaction business 1
03 and the second computer 200, the failure notification means 1
03 received a transaction business switching instruction,
It is possible to import all external devices such as databases and journals including databases and journals that are required to execute the transaction work that was executed in the active system, roll back the transaction work, and continue executing the transaction work. Transaction business switching means 202 for recovering the state, transaction business alternative processing means 201 for succeeding and executing subsequent transaction business after recovery of the transaction business, and continuing operation, and batch business for executing batch business during normal operation The external storage device 500 monitors the work load of the processing unit 203 and the second computer 200 at regular monitoring time intervals.
The first work load information monitoring means 204 which records the work load information of the second computer 200 in the work load information table 501 created in the above, and the execution of the transaction work due to the failure in the first computer 100 are switched. When the transaction work is resumed by the transaction work alternative processing means 201, the first work load information monitoring means 20
4 acquires the work load information of the second computer 200 and the work load information of the third computer 300, and the second computer 20
Batch work switching that detects that the work load of 0 is high, further identifies the one with a low work load on the third computer 300, and determines whether to switch the batch work to the third computer 300 When it is determined that the batch operation is switched to the determination unit 205 and the third computer 300,
A batch job switching notification unit 206 for instructing the third computer 300 to switch the batch job, and a first batch job switching processing unit 30 for switching the batch job from the second computer 200 in the third computer 300.
2 and a first batch job alternative processing unit 301 that takes over the switched batch job and continues the subsequent execution.

In this way, when the first computer 100 becomes a failure, when the transaction work being operated on the first computer 100 is switched to the second computer 200, this causes By preventing the work load of the second computer 200 from becoming extremely higher than that of the third computer 300, and switching the batch work that was being operated by the second computer 200 to the third computer 300 and operating it, It is possible to equalize the work load of.

[0026]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[First Embodiment] Referring to FIG.
In the hot standby system switching control method according to the first embodiment of the present invention, there is a failure in the first computer 100 in which the active system that executes the transaction business is present during normal business operation, and in the first computer 100. A second computer 200 having an alternative system that takes over the transaction work from the active system and continues execution when the transaction work cannot be executed in the active system
A third computer 300 that does not include an active system and an alternative system,
From the external storage device 500 that is channel-connected from the first computer 100, the second computer 200, and the third computer 300, and is shared by all the computers,
Its main part is composed.

The first computer 100 is a transaction business processing means 101 for executing transaction business during normal business operation in an active system, and the first computer 10.
When a failure occurs in 0, the failure event determination means 102 for determining whether or not the transaction business operation in the active system can be continued, and the failure event determination means 102 make it impossible to continue the transaction business in the active system. In order to switch the execution of the transaction work to the alternative system existing in the second computer 200 when it is determined, the second computer 200
The failure notification means 103 for notifying a failure to the transaction work switching means 202 is included.

The second computer 200 has a fault notifying means 10
3 is notified of the occurrence of a failure, receives an instruction to switch transactional work, imports all the disk devices that are running in the active system, including the databases and journals required for executing transactional work, and when a failure occurs The transaction transaction switching unit 202 that rolls back the transaction that has been executed and recovers the subsequent transaction transaction to a state in which the transaction transaction can be continued, and the transaction transaction switching unit 202 that is used when the transaction transaction switching process ends. Transaction business alternative processing means 201 for taking over the operation of transaction business and batch business processing means 2 for executing batch business during normal business operation
03 and the work load of the second computer 200 are monitored at regular monitoring time intervals and recorded as work load information of the second computer 200 in the work load information table 501 created in the external storage device 500. , The first computer 100
When a transaction has been switched to the alternative system of the second computer 200 due to a failure in the second computer 200, the workload of the second computer 200 at the time after the transaction is switched by referring to the workload information table 501. Information and a work load information of the third computer 300, the first work load information monitoring means 204, and the second computer 20.
A batch that compares the work load information of 0 and the work load information of the third computer 300, and determines that the batch work is to be switched to the third computer 300 when the load of the third computer 300 is lower. Business switching determination means 205,
When the batch job is switched to the third computer 300, the batch job switching notifying unit 206 for notifying the third computer 300 of the switching of the batch job is included.

The third computer 300 is the third computer 30.
A second work load information monitoring unit 303 which monitors the work load of 0 at a constant monitoring time interval and records the work load information of the third computer 300 in the work load information table 501 created in the external storage device 500. And a first batch job switching processing means 302 for switching the batch job to the third computer 300 when the batch job switching notifying means 206 of the second computer 200 notifies the batch job switching. It is configured to include a first batch job alternative processing unit 301 that continues execution of the batch job after taking over the batch job switched by the first batch job switching processing unit 302.

The external storage device 500 is the second computer 20.
The work load information table 501 records the work load information of 0 and the work load information of the third computer 300 in association with the information collection time.

Each of these means operates roughly as follows.

The transaction business processing means 101 is
Execute transaction business during normal business operation.

The fault event judging means 102 detects a fault when a fault occurs in the main body of the first computer 100, and
It is determined whether or not it is possible to continue the operation of the transaction work on the computer 100, and when it is determined that the operation cannot be continued, the failure notification unit 103 is notified.

When the failure notification means 103 receives the notification from the failure event determination means 102, the transaction work switching means 20 of the second computer 200 in which the alternate system exists.
Instruct 2 to switch the transaction work to the alternative system.

Transaction work switching means 202
Receives a transaction business switching instruction from the failure notifying means 103, rolls back the transaction being executed by the transaction business processing means 101 of the first computer 100, and resumes the operation of the subsequent transaction business. To

Transaction business alternative processing means 201
Receives the completion of the recovery of the transaction work by the transaction work switching means 202 and restarts the subsequent transaction work.

The batch job processing means 203 executes a batch job in the second computer 200 during normal business operation.

The first work load information monitoring means 204 monitors the work load of the second computer 200 at a constant monitoring time interval, and stores the second work load information table 501 in the external storage device 500 in the second work computer 200. If the transaction work is switched to the alternative system of the second computer 200 due to a failure in the first computer 100, the work load information table 501
Refer to the workload information of the second computer 200 and the third
The work load information of the computer 300 is acquired.

When the transaction work is switched to the alternative system, the batch work switching judging means 205 and the third computer 300 work load information of the second computer 200.
If the work load of the third computer 300 is lower than the work load information of the third computer 300, the batch work is compared with the third computer 30.
Determine to switch to 0.

When the batch work switching judging means 205 judges that the batch work should be switched to the third computer 300, the batch work switching notifying means 206 gives an instruction to switch the batch work to the third computer 300.

First batch job switching processing means 302
Performs a batch job switching process when the batch job switching notification unit 206 of the second computer 200 instructs the batch job switching.

The first batch job alternative processing means 301 is
The batch job switched to the third computer 300 is continued to execute the batch job after the succeeding batch job.

The second work load information monitoring means 303, like the first work load information monitoring means 204, monitors the work load of the third computer 300 at regular monitoring time intervals and creates it in the external storage device 500. Workload information table 50
1 is recorded as the work load information of the third computer 300,
In preparation for the reference of the work load information by the first work load information monitoring means 204 when the transaction work is switched to the alternative system.

FIG. 2 shows the contents of the work load information table 501 created in the external storage device 500. At each of times t1, t2, t3, t4, ...
The work load information of each computer 200 is p21, p22, p2 by the first work load information monitoring means 204, respectively.
3, p24, ..., The third computer 300 is recorded.
The work load information of the second work load information monitoring unit 303
, P31, p32, p33, p34, ...
Is recorded. For example, the work load information of the second computer 200, which is acquired by the first work load information monitoring unit 204 at the time t4 when the transaction work switching process ends, is p24, and the work load information of the third computer 300. Is p34.

FIG. 3 (a) shows the second operation during business operation.
CPU (Central Process) of the computer 200 of
The processing of the first work load information monitoring unit 204 that manages the load status of the (Singing Unit) as work load information is shown. This process is performed by the timer setting step A1.
And a work load information acquisition step A2 and a work load information recording step A3.

FIG. 3 (b) shows the third example during the business operation.
The processing of the second work load information monitoring means 303 that manages the load status of the CPU of the computer 300 as work load information is shown. This process is the timer setting step A4.
And a work load information acquisition step A5 and a work load information recording step A6.

FIG. 4 shows the switching processing of transaction work when a failure occurs in the first computer 100. This process includes a failure detection step B1, an operational transaction work continuability determination step B2,
Failure notification step B3 and switching instruction step B4
Switching process step B5, switching process end notification step B6, and switching process end notification step B7
Consists of.

FIG. 5 shows the first work load information monitoring means 2 when a failure occurs in the first computer 100 and the transaction work is switched to the alternative system of the second computer 200.
Reference numeral 04 denotes the batch job switching process executed by the second computer 200 when the fact that the transaction job switching process has been completed is notified. This process includes a work load information acquisition step C1, a work load information passing step C2, a batch work switching necessity determination step C3, a batch work switching notification step C4,
It comprises a batch job switching instruction step C5 and a batch job switching step C6.

Next, the operation of the hot standby system switching control system according to the first embodiment configured as described above will be described in detail with reference to FIGS. 1, 2, 3, 4, and 5. explain.

First, the first work load information monitoring means 204
Sets a timer to set a constant monitoring time interval for acquiring the work load information of the second computer 200 (step A1). Next, when the timer in step A1 times out, the first work load information monitoring means 204
The work load information of the second computer 200 at that time is acquired (step A2). Further, the first work load information monitoring means 204 stores the acquired work load information and acquisition time information of the second computer 200 in the external storage device 500.
It is recorded in the work load information table 501 created in step S3 (step A3).

On the other hand, the second work load information monitoring means 303
Is similar to the first work load information monitoring means 204 and is the third
A timer is set to set a constant monitoring time interval for acquiring the work load information of the computer 300 (step A)
4). Next, when the timer of step A4 times out, the second work load information monitoring means 303 acquires the work load information of the third computer 300 at that time (step A5). Further, the second work load information monitoring unit 303 records the acquired work load information and acquisition time information of the third computer 300 in the work load information table 501 created in the external storage device 500 (step A6). .

When a failure occurs in the first computer 100,
The failure event determination means 102 detects a failure (step B
1), it is determined whether or not the transaction work running in the active system of the first computer 100 can be continuously operated on the computer (step B2), and the first computer 100
If it is determined that it is impossible to continue the operation in step S1, the failure notification means 103 is notified that a failure has occurred (step B
3). Next, the failure notification means 103 that has received the notification
The transaction business switching means 202 of the second computer 200 is instructed to switch the transaction business (step B4).

The transaction work switching means 202 which has received the transaction work switching instruction
The transaction that was being executed at the time when the failure occurred in the computer 100 is rollback-restored, the transaction transaction is re-executed, and a switching process is executed to take over and execute the subsequent transaction transaction (step B5). When the switching process ends, the transaction work switching unit 202 notifies the transaction work alternative processing unit 201 that the switching process has ended (step B6), and the first work load information monitoring unit 2
04 is notified that the switching process is completed (step B7). The transaction work alternative processing means 201, which has been notified that the switching process has been completed, takes over the subsequent transaction work operation in the alternative system.

In the switching of the batch work, first, the first work load information monitoring means 204 uses the work load information of the second computer 200 and the The work load information of the computer 300 of No. 3 is acquired (step C1), and the work load information is passed to the batch work switching determination means 205 (step C2).

The batch work switching determination means 205 refers to the work load information, and the work load of the second computer 200 is higher than the work load of the third computer 300, and the subsequent work operation causes delay or performance degradation. It is determined whether or not there is a possibility (Step C3). Second computer 2
When it is determined that the work load of 00 is high and there is a possibility that the subsequent work operation may cause a delay or a performance degradation, the batch work switch determination unit 205 recognizes that the batch work needs to be switched, and switches the batch work. The notification means 206 is notified (step C4).

Receiving the notice, the batch job switching notifying means 206 switches the batch job by the third computer 3.
00 to the first batch job switching processing means 302 (step C5).

First batch job switching processing means 302
When receiving the instruction to switch the batch job, the CPU performs a switching process for handing over the batch job being executed by the batch job processing means 203 to the first batch job substitute processing means 301 (step C6).

When the switching process is completed, the first batch job alternative processing means 301 restarts the batch job and succeeds and executes the subsequent batch jobs.

As described above, according to the first embodiment,
In addition to a computer with an active system and a computer with an alternative system,
In a hot standby system with one computer connected, if a computer with an active system fails and it becomes impossible to continue transaction business operations in the active system, the transaction in the alternate system When the business load of a computer with an alternative system becomes high by executing the transaction recovery and continuing the operation of the transaction business and switching and executing the transaction business By switching the executed batch operation to another computer and executing the same, it is possible to equalize the work load of the work in each computer in the loosely coupled multiprocessor system and prevent the work delay and performance deterioration. The reason is that the CP with the alternative computer and the other computer
If the workload of the computer with the alternate system becomes high when the transaction workload is monitored, such as the U load status, and the transaction workload is switched to the alternate system, the batch jobs executed on the computer with the alternate system will be executed. This is because it is executed by switching to another computer.

[Second Embodiment] Referring to FIG. 6,
In the hot standby system switching control method according to the second embodiment of the present invention, there is a failure in the first computer 100 in which the active system that executes transaction business exists during normal business operation, and in the first computer 100. Occurs,
A second computer 200 having an alternative system that takes over the transaction work from the active system and continues execution when the transaction work cannot be executed in the active system;
A third computer 300 and a fourth computer 400, which are two computers that do not include an active system and a substitute system, and a first computer 100, a second computer 200, a third computer 300, and a fourth computer. The external storage device 500 is connected to each of the computers 400 and shared by all computers.

The first computer 100 is a transaction business processing means 101 for executing transaction business during normal business operation in an active system, and the first computer 10.
When a failure occurs in 0, the failure event determination means 102 for determining whether or not the transaction business operation in the active system can be continued, and the failure event determination means 102 make it impossible to continue the transaction business in the active system. In order to switch the execution of the transaction work to the alternative system existing in the second computer 200 when it is determined, the second computer 200
The failure notification means 103 for notifying a failure to the transaction work switching means 202 is included.

The second computer 200 has the fault notifying means 10
3 is notified of the occurrence of a failure, receives a transaction operation switching instruction, imports all the disk devices including the databases and journals that are required to execute the transaction operation executed in the active system, and executes them when an error occurs. Transaction business switching means 20 that rolls back and recovers the existing transaction to a state in which the execution of subsequent transaction business can be continued.
2, and when the transaction work switching unit 202 completes the transaction work switching process, the transaction work alternative processing unit 201 that takes over the operation of the subsequent transaction work, and the batch work process that executes the batch work during normal business operation Means 203
And the work load of the second computer 200 is monitored at regular monitoring time intervals and recorded in the work load information table 501 created in the external storage device 500 as the work load information of the second computer 200. When a failure occurs in the first computer 100 and the transaction work is switched to the alternative system of the second computer 200, the work load information table 501 is referred to, and the second time is set after the transaction work is switched. Workload information of computer 200, workload information of third computer 300 and fourth computer 4
First work load information monitoring means 204 for acquiring work load information of 00, work load information of the second computer 200, work load information of the third computer 300, and fourth work computer 4
00, and when the load of the third computer 300 or the fourth computer 400 is lower than that of the second computer 200, the batch work is compared to the third computer 300.
Alternatively, the batch job switching determination unit 205 that determines to switch to the fourth computer 400 and the third computer 300 or the fourth computer 400 when switching the batch job to the third computer 300 or the fourth computer 400. It is configured to include a batch job switching notification unit 206 for notifying switching of batch jobs.

The third computer 300 is the third computer 30.
A second work load information monitoring means 303 for monitoring the work load of 0 at constant monitoring time intervals and recording the work load information in the work load information table 501 created in the external storage device 500; and a second computer. When the batch job switching notifying means 206 of 200 notifies the batch job switching, the first batch job switching processing means 302 for performing the batch job switching processing to the third computer 300, and the first batch job switching It is configured to include a first batch job alternative processing means 301 for continuing the operation of the batch job after taking over the batch job switched by the processing means 302.

The fourth computer 400 is the third computer 30.
Similarly to 0, the third work load information in which the work load of the fourth computer 400 is monitored at constant monitoring time intervals and the work load information is recorded in the work load information table 501 created in the external storage device 500. When the monitoring means 403 and the batch work switching notifying means 206 of the second computer 200 are notified of the batch work switching, the fourth computer 400
Second batch job switching processing unit 402 that performs batch job switching processing to the second batch job, and a second batch that continues operation of the batch job that has been switched over by the second batch job switching processing unit 402 The work substitute processing means 401 is included.

The external storage device 500 is the second computer 20.
A work load information table 5 in which the work load information of 0, the work load information of the third computer 300, and the work load information of the fourth computer 400 are recorded in association with the information collection time.
01 is included.

Each of these means has the following functions.

The transaction business processing means 101 is
Execute transaction business during normal business operation.

When a failure occurs in the first computer 100 main body, the failure event determination means 102 detects the failure and
It is determined whether or not it is possible to continue the operation of the transaction work on the computer 100, and when it is determined that the operation cannot be continued, the failure notification unit 103 is notified.

When the failure notification means 103 receives the notification from the failure event determination means 102, the transaction work switching means 20 of the second computer 200 in which the alternate system exists.
Instruct 2 to switch the transaction work to the alternative system.

Transaction business switching means 202
Can receive an instruction for switching transactional work from the failure notifying unit 103, roll back the transaction in which the transactional work of the first computer 100 was being executed, and resume the operation of the subsequent transactional work. To do so.

Transaction business alternative processing means 201
Receives the completion of the recovery of the transaction work by the transaction work switching means 202 and restarts the subsequent transaction work.

The batch job processing means 203 executes the batch job in the second computer 200 during normal business operation.

The first work load information monitoring means 204 monitors the work load of the second computer 200 at regular monitoring time intervals and records it as work load information in the work load information table 501 created in the external storage device 500. In addition, if a failure occurs in the first computer 100 and the second computer 2
When the transaction work is switched to the alternative system of 00, the work load information table 501 is referred to, the work load information of the second computer 200, the work load information of the third computer 300, and the work load information of the fourth computer 400. To get.

When the transaction work is switched to the alternative system, the batch work switching determination means 205 determines the work load information of the second computer 200 and the third computer 300.
Of the workload information of the fourth computer 400 and the workload information of the fourth computer 400, and the second computer 200 to the third computer 30
If the load of 0 or the fourth computer 400 is lower,
It is determined to switch the batch job to the third computer 300 or the fourth computer 400.

When the batch work switching judging means 205 judges that the batch work should be switched to the third computer 300, the batch work switching notifying means 206 notifies the third computer 300 of the batch work switching, When it is determined to switch the batch job to the fourth computer 400, the fourth computer 400 is notified of the batch job switch.

First batch job switching processing means 302
Performs the batch job switching process when notified of the batch job switching from the batch job switching notifying unit 206 of the second computer 200.

The first batch job alternative processing means 301 is
The batch job switched to the third computer 300 is continuously taken over and the subsequent batch jobs are executed.

The second work load information monitoring means 303, like the first work load information monitoring means 204, monitors the work load of the third computer 300 at regular monitoring time intervals and creates it in the external storage device 500. Workload information table 50
The work load information is recorded in No. 1 to prepare for reference of the work load information by the first work load information monitoring means 204 when the transaction work is switched to the alternative system.

Second batch job switching processing means 402
Similarly to the first batch job switching processing unit 302, when the batch job switching notifying unit 206 of the second computer 200 notifies the batch job switching, the batch job switching process is performed.

The second batch job alternative processing means 401 is
The batch job after switching to the fourth computer 400 is continuously executed.

The third work load information monitoring means 403, like the first work load information monitoring means 204 and the second work load information monitoring means 303, monitors the work load of the fourth computer 400 for a certain monitoring time. External storage device 50 monitored at intervals
The work load information of the fourth computer 400 is recorded in the work load information table 501 created in 0, and the reference of the work load information by the first work load information monitoring means 204 when the transaction work is switched to the alternative system Prepare for

FIG. 7 shows the contents of the work load information table 501 created in the external storage device 500. At each of times t1, t2, t3, t4, ...
The work load information of each computer 200 is p21, p22, p2 by the first work load information monitoring means 204, respectively.
3, p24, ..., The third computer 300 is recorded.
The work load information of the second work load information monitoring unit 303
, P31, p32, p33, p34, ...
The work load information of the fourth computer is p4 by the third work load information monitoring means 403.
It is recorded as 1, p42, p43, p44, .... For example, the work load information of the second computer 200, which is acquired by the first work load information monitoring unit 204 at the time t4 when the transaction work switching process ends, is p24, and the work load information of the third computer 300. Is p34, and the work load information of the fourth computer 400 is p44.

FIG. 8A shows the second status during the business operation.
The processing of the first work load information monitoring means 204 for managing the load status of the CPU of the computer 200 as work load information is shown. This process is performed by the timer setting step D1.
And a work load information acquisition step D2 and a work load information recording step D3.

FIG. 8 (b) shows the third status during business operation.
The processing of the second work load information monitoring means 303 that manages the load status of the CPU of the computer 300 as work load information is shown. This process is the timer setting step D4.
And a work load information acquisition step D5 and a work load information recording step D6.

FIG. 8C shows the fourth line during the business operation.
The processing of the third work load information monitoring means 403 for managing the load status of the CPU of the computer 400 as work load information is shown. This process is performed by the timer setting step D7.
And a work load information acquisition step D8 and a work load information recording step D9.

FIG. 9 shows the first work load information monitoring means 2 when the transaction work is switched to the alternative system of the second computer 200 due to the failure of the first computer 100.
Reference numeral 04 indicates the batch job switching processing executed by the second computer 200 when the fact that the transaction job switching processing has been completed is notified. This process includes a work load information acquisition step E1, a work load information passing step E2, a batch work switching necessity determination step E3, a batch work switching notification step E4,
It comprises a batch job switching instruction step E5 and a batch job switching step E6.

Next, the operation of the hot standby system switching control system according to the second embodiment configured as described above will be described in detail with reference to FIGS. 6, 7, 8 and 9.

First, the first work load information monitoring means 204
Sets a timer to set a constant monitoring time interval for acquiring the work load information of the second computer 200 (step D1). Next, when the timer of step D1 times out, the first work load information monitoring means 204
The work load information of the second computer 200 at that time is acquired (step D2). Further, the first work load information monitoring means 204 stores the acquired work load information and acquisition time information of the second computer 200 in the external storage device 500.
It is recorded in the work load information table 501 created in step S3 (step D3).

On the other hand, the second work load information monitoring means 303
Is similar to the first work load information monitoring means 204 and is the third
A timer to set a constant monitoring time interval for acquiring the workload information of the computer 300 (step D)
4). Next, when the timer of step D4 times out, the second work load information monitoring means 303 acquires the work load information of the third computer 300 at that time (step D5). Further, the second work load information monitoring means 303 records the acquired work load information and acquisition time information of the third computer 300 in the work load information table 501 created in the external storage device 500 (step D6). .

The third work load information monitoring means 403
Is the same as the first work load information monitoring unit 204 and the second work load information monitoring unit 303, the fourth computer 40.
A timer is set in order to set a constant monitoring time interval for acquiring the work load information of 0 (step D7). next,
When the timer in step D7 times out, the third work load information monitoring means 403 causes the third work load information monitoring means 403 to perform the fourth operation at that time.
The workload information of the computer 400 of the above (Step D
8). Furthermore, the third work load information monitoring means 403
The acquired work load information and acquisition time information of the fourth computer 400 are recorded in the work load information table 501 created in the external storage device 500 (step D9).

The operation of switching the transaction work when a failure occurs in the first computer 100 is the same as that of the first embodiment, and the first operation described with reference to FIG.
The detailed description of the operation will be omitted after the description of the operation in the embodiment.

In the switching of batch work, first, the first work load information monitoring means 204 uses the work load information table 501 to check the work load information of the second computer 200 at the time when the transaction work switching process is completed. Three
Workload information of the computer 300 and the fourth computer 40
The work load information of 0 is acquired (step E1), and the work load information is passed to the batch work switching determination means 205 (step E2).

The batch work switching determination means 205 refers to the work load information, and the work load of the second computer 200 is higher than the work load of the third computer 300 or the work load of the fourth computer 400. Then, it is determined whether or not the subsequent business operation may cause delay or performance degradation (step E3). Furthermore, when the batch work switching determination unit 205 determines that the work load of the second computer 200 is high and there is a possibility that the subsequent work operation will cause delay or performance degradation, the third computer 300
And the workload of the fourth computer 400 are compared, and the batch task switching notifying means 206 is notified that the batch task is switched to the computer with the lower task load (step E4).

Upon receiving the notification, the batch job switching notifying means 206 instructs the first batch job switching processing means 302 of the third computer 300 to switch the batch job when switching the batch job to the third computer 300. Then, when the batch job is switched to the fourth computer 400, the second batch job switching processing unit 402 of the fourth computer 400 is instructed to switch the batch job (step E5).

First batch job switching processing means 302
Alternatively, when the second batch job switching processing means 402 receives the switching instruction, the batch job processing means 20
The first batch job alternative processing means 301 or second batch job alternative processing means 4
The switching process to carry over to 01 is performed (step E
6).

When the switching process is completed, the first batch job alternative processing unit 301 or the second batch job alternative processing unit 401 restarts the batch job and succeeds and executes the subsequent batch jobs.

Also in the second embodiment, when the first computer 100 becomes a failure, the transaction work operated on the first computer 100 is changed to the second computer 20.
This causes the second computer 2 to
It is possible to prevent the work load of 00 from becoming extremely higher than that of the third computer 300 and the fourth computer 400, and to execute the batch work that has been operated on the second computer 200 by the third computer 30.
By switching to 0 or the fourth computer 400 for operation, it is possible to equalize the work load of each computer.

[Third Embodiment] FIG. 10 is a block diagram showing the configuration of a hot standby system switching control system according to a third embodiment of the present invention. Referring to FIG. 10, in the third exemplary embodiment of the present invention, the first computer 110 in the first exemplary embodiment shown in FIG. In contrast, the second program 210 is different from the third computer 300 in that the third program 310 is provided.

The first program 110 is read by the first computer 100 and controls the operation of the first computer 100 as the transaction work processing means 101, the failure event determination means 102 and the failure notification means 103. First
The operation of the first computer 100 under the control of the program 110 of the above is the first computer 100 in the first embodiment.
Since it is exactly the same as the operation of, the detailed explanation is omitted.

Further, the second program 210 is read by the second computer 200, and the operation of the second computer 200 is changed to the transaction work alternative processing means 201, the transaction work switching means 202, the batch work processing means 203, The work load information monitoring unit 204, the batch work switching determination unit 205, and the batch work switching notifying unit 206 are controlled. Second program 21
The operation of the second computer 200 under the control of 0 is exactly the same as the operation of the second computer 200 in the first embodiment, and therefore its detailed description is omitted.

Further, the third program 310 is
The third computer 300 reads the data and controls the operation of the third computer 300 as the first batch work substitute processing means 301, the first batch work switching processing means 302, and the second work load information monitoring means 303. The operation of the third computer 300 under the control of the third program 310 is
Since the operation is the same as that of the third computer 300 in the embodiment, detailed description thereof will be omitted.

[Fourth Embodiment] FIG. 11 is a block diagram showing the configuration of a hot standby system switching control system according to a fourth embodiment of the present invention. Referring to FIG. 11, in the fourth exemplary embodiment of the present invention, the first program 110 is stored in the second computer 200 for the first computer 100 in the second exemplary embodiment shown in FIG. On the other hand, the second program 210 is used for the third computer 300 and the fourth computer 400 and the third program 31
They are different in that each has 0.

The first program 110 is read by the first computer 100 and controls the operation of the first computer 100 as the transaction work processing means 101, the failure event determination means 102 and the failure notification means 103. First
The operation of the first computer 100 under the control of the program 110 of the first embodiment is the same as that of the first computer 100 in the second embodiment.
Since it is exactly the same as the operation of, the detailed explanation is omitted.

The second program 210 is read by the second computer 200, and the operation of the second computer 200 is changed to the transaction work alternative processing means 201, transaction work switching means 202, batch work processing means 203, The work load information monitoring unit 204, the batch work switching determination unit 205, and the batch work switching notifying unit 206 are controlled. Second program 21
The operation of the second computer 200 under the control of 0 is exactly the same as the operation of the second computer 200 in the second embodiment, and therefore its detailed description is omitted.

Further, the third program 310 is
Read by the third computer 300 and the fourth computer 400, and the operations of the third computer 300 and the fourth computer 400 are performed by the first batch job substitution processing means 301, the first batch job switching processing means 302 and the second batch job switching processing means 302 and the second batch job switching processing means 302. It is controlled as the work load information monitoring means 303. The operations of the third computer 300 and the fourth computer 400 under the control of the third program 310 are exactly the same as the operations of the third computer 300 and the fourth computer 400 in the second embodiment. I will omit the detailed explanation.

[0107]

The first effect is that in a hot standby system having a configuration in which one or more computers are connected in addition to a computer having an active system and a computer having an alternative system, a computer having an active system is provided. When a failure occurs and it becomes impossible to continue the operation of the transactional business in the active system, the transactional business is restored in the alternate system to continue the operation of the transactional business, and the transactional business is switched to the alternative system. When the workload of the computer with the alternate system becomes high by executing the above, the batch tasks being executed on the computer with the alternate system are switched to another computer and executed Regarding the system, it is possible to equalize the work load of work on each computer and prevent work delay and performance deterioration. The reason is that the computer having the alternative system and the other computer monitor the state of the business load such as the CPU load status, and when the transaction work is switched to the alternative system, the business load of the computer having the alternative system becomes high. This is because, in the case of executing, the batch work being executed by the computer having the alternative system is switched to another computer for execution.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a hot standby system switching control method according to a first embodiment of the present invention.

FIG. 2 is a diagram showing the contents of a work load information table in FIG.

FIG. 3 is a flowchart showing a process of a first work load information monitoring means and a second work load information monitoring means in FIG.

FIG. 4 is a flow chart showing processing of switching transaction work when a failure occurs in the first computer in FIG.

FIG. 5 is a flow chart showing a process of switching a batch job executed by the second computer in FIG.

FIG. 6 is a block diagram showing a configuration of a hot standby system switching control method according to a second embodiment of the present invention.

7 is a diagram showing the contents of a work load information table in FIG.

8 is a flow chart showing the processing of a first work load information monitoring means, a second work load information monitoring means and a third work load information monitoring means in FIG.

FIG. 9 is a flow chart showing a process of switching a batch job, which is executed by the second computer in FIG.

FIG. 10 is a block diagram showing a configuration of a hot standby system switching control method according to a third embodiment of the present invention.

FIG. 11 is a block diagram showing a configuration of a hot standby system switching control method according to a fourth embodiment of the present invention.

FIG. 12 is a block diagram showing a configuration of a conventional hot standby system.

[Explanation of symbols]

100 first calculator 101 transaction business processing means 102 Failure event determination means 103 Failure notification means 200 Second Calculator 201 Transaction business alternative processing means 202 Transaction business switching means 203 Batch business processing means 204 First work load information monitoring means 205 Batch job switching determination means 206 Batch job switching notification means 300 Third Calculator 301 First batch job alternative processing means 302 First batch job switching processing means 303 Second work load information monitoring means 400 Fourth Calculator 401 Second batch job alternative processing means 402 Second batch job switching processing means 403 Third work load information monitoring means 500 external storage device 501 workload information table

Claims (12)

[Claims] 1. A loosely coupled multiprocessor system in which three computers are connected in a channel, and a first computer having an active system for executing transaction business during normal business operation and a first computer. If there is a failure that occurred in the other computer and the transaction business in the active system cannot be continued, the failed transaction business is restored and the operation of the subsequent transaction business is taken over and executed. In the hot standby system having a configuration in which the second computer and the third computer are connected, the first computer executes transaction business as an active system during normal business operation, and transaction business processing means, If a failure occurs on another computer, determine whether the transaction business in the active system can be continued When the adverse event determination means and the failure event determination means determine that the transaction business in the active system cannot be continued, the transaction business switching means existing in the second computer is notified of the failure and the transaction business is switched. A transaction notifying means for instructing the transaction transaction, and the second computer receives the instruction to switch the transaction transaction from the failure notifying means and restores the transaction transaction being executed in the active system to a state where execution can be continued. A switching means, a transaction business alternative processing means for continuing and operating the subsequent transaction business after recovery of the transaction business, a batch business processing means for executing the batch business during normal operation, and a second computer Second calculation by monitoring the work load at regular monitoring time intervals First work load information monitoring means for recording the work load information of the above in the work load information table, and execution of the transaction work is switched due to a failure in the first computer, and the transaction work is resumed by the transaction work alternative processing means. In this case, the workload information of the second computer is compared with the workload information of the third computer by referring to the workload information table, and it is detected that the workload of the second computer has increased. And a batch job switching judging means for judging whether to switch the batch job to the third computer.
If the third computer is determined to switch the batch job, the third computer is instructed to switch the batch job to the third computer.
The work load of the computer is monitored at regular monitoring time intervals.
Second work load information monitoring means for recording work load information of another computer in the work load information table, and batch work switching processing for switching the batch work instructed to be switched by the batch work switching notifying means from the second computer A hot standby system switching control method comprising: means and a batch job alternative processing means for continuing execution of the batch job switched by the batch job switching processing means. 2. The workload information table of the second computer and the workload information of the third computer are recorded in the workload information table in correspondence with each time. Hot standby system switching control method. 3. A loosely-coupled multiprocessor system in which four or more computers are connected in a channel, and during normal business operation, a first computer having an active system for executing transaction business and a first computer, If a failure occurred on the computer of No. 1 makes it impossible to continue the transactional business in the active system, it has an alternative system that recovers the failed transactional business and takes over the operation of the subsequent transactional business. In a hot standby system having a configuration in which a second computer and two or more third computers are connected, a transaction business process in which the first computer executes transaction business as an active system during normal business operation If the means and the first computer fail, is it possible to continue the transaction work in the active system? A fault event determining means for determining whether,
Failure notification means for notifying the transaction operation switching means existing in the second computer of the failure and instructing the transaction operation switching when the failure event determination means determines that the transaction operation in the active system cannot be continued And a transaction operation switching unit for recovering the transaction operation executed in the active system to a state in which the second computer receives the instruction to switch the transaction operation from the failure notifying unit, and the transaction operation After the recovery of the transaction, the transaction transaction alternative processing means for taking over and executing the subsequent transaction transactions and continuing the operation, the batch transaction processing means for executing the batch transaction during normal operation, and the constant monitoring of the workload of the second computer Monitor the workload information of the second computer by monitoring at time intervals. The first work load information monitoring means for recording in the load information table and the work load when the execution of the transaction work is switched due to a failure in the first computer and the transaction work is resumed by the transaction work alternative processing means. The work load information of the second computer is compared with the work load information of the third computer by referring to the information table, and when it is detected that the work load of the second computer has increased, the third computer In the case where it is determined that the batch work is switched to the third computer, the batch work switching determination means that determines whether the work load is low in the third computer and determines whether to switch the batch work to the third computer. And a batch job switching notification means for instructing the third computer to switch the batch job.
A second work load information monitoring means for monitoring the work load of the third computer at a constant monitoring time interval and recording the work load information of the third computer in the work load information table; and the batch work. A batch job switching processing means for switching a batch job instructed to be switched by the switching notifying means from the second computer, and a batch job alternative processing means for continuing execution of the batch job switched by the batch job switching processor A hot-standby system switching control method comprising: 4. The work load information table records work load information of a second computer and work load information of two or more third computers in correspondence with each time. The hot standby system switching control method according to claim 3. 5. A loosely coupled multiprocessor system in which three computers are connected in a channel, and a first computer having an active system for executing transaction business during normal business operation and a first computer. If there is an alternative computer that has a failure in the operating system that makes it impossible to continue the transaction business, the failed transaction business is restored and the subsequent transaction business is taken over and executed. In a hot standby system with a configuration in which the second computer and the third computer are connected, a failure occurs in the first computer and it becomes impossible to continue the transaction work on the first computer. In the second computer, the transaction business is restored and the operation of the transaction business is taken over from the first computer. The hot standby system switching control is characterized in that when the work load of the second computer becomes high by this, the batch work being executed in the second computer is switched to the third computer and executed. Method. 6. A loosely-coupled multiprocessor system in which four or more computers are connected in a channel, and during normal business operation, a first computer having an operational system for executing transaction business and a first computer, If a transaction that occurs in the active system makes it impossible to continue transactional transactions in the active system, the failed transactional transaction is restored and the subsequent transactional transaction is taken over and executed. In a hot standby system having a configuration in which a second computer and two or more third computers are connected, a failure occurs in the first computer and continuation of transaction work on the first computer is not possible. When it becomes possible, the transaction is restored on the second computer, and the operation of the transaction is performed by the first calculation. When the business load of the second computer becomes high, the third computer identifies the one with the low business load, and the third computer executes the second computer. A hot standby system switching control method characterized in that batch jobs are switched and executed. 7. A transaction business processing means for executing transaction business as an active system during normal business operation of a computer, and whether or not the transaction business can be continued in the active system when a failure occurs in the computer. Failure transaction determining means, and when the failure event determining means determines that the transaction business in the active system cannot be continued, the transaction business is switched to the transaction business switching means existing in another computer, and the transaction business is performed. A program for operating as failure notification means for instructing switching of the. 8. A transaction business switching means for recovering a computer to a state in which it is possible to continue execution of a transaction business executed in an active system in response to a transaction business switching instruction from a failure notification means of another computer Alternative transaction processing means that takes over and executes the subsequent transaction operations after recovery, continues the operation, batch job processing means that executes batch operations during normal operation, and monitors the workload of the computer at regular monitoring time intervals Then, the first work load information monitoring means for recording the work load information of the computer in the work load information table, the execution of the transaction work is switched due to a failure in the first computer, and the transaction work is executed by the transaction work alternative processing means. Was resumed In this case, the workload information of the computer of the computer is compared with the workload information of another computer by referring to the workload information table, and when it is detected that the workload of the computer becomes high, another A batch job switching determination means for determining whether to switch the batch job to a computer, and a batch job for instructing the other computer to switch the batch job when it is determined to switch the batch job to the other computer A program for operating as switching notification means. 9. A transaction business switching means for recovering a computer to a state in which it is possible to continue execution of a transaction business executed in an active system in response to a transaction business switching instruction from a failure notification means of another computer Alternative transaction processing means that takes over and executes the subsequent transaction operations after recovery, continues the operation, batch job processing means that executes batch operations during normal operation, and monitors the workload of the computer at regular monitoring time intervals Then, the first work load information monitoring unit that records the work load information of the computer in the work load information table, the execution of the transaction work is switched due to a failure in the first computer, and the transaction work is resumed by the transaction work alternative processing unit. If done The workload information of the computer of the computer is compared with the workload information of another computer by referring to the workload information table, and when it is detected that the workload of the computer becomes high, the other computer Batch job switching determination means that determines the one with a low business load and determines whether to switch the batch job to the other computer, and batch job when it is determined to switch the batch job to the other computer A program for operating as a batch job switching notification means for instructing the other computer to switch. 10. A second method for monitoring a computer work load of the computer at a constant monitoring time interval and recording the work load information of the computer in a work load information table.
Workload information monitoring means, a batch job switching processing means for switching a batch job instructed to be switched by another computer from the other computer, and a batch job switched by the batch job switching processing means for execution after handing over. A program for operating as a continuous batch job alternative processing means. 11. A transaction business processing means for executing transaction business as an active system during normal business operation of a computer, and whether or not the transaction business can be continued in the active system when a failure occurs in the computer. Failure event determining means for determining whether the transaction operation switching means existing in another computer is notified of the failure when the failure event determining means determines that the transaction operation in the active system cannot be continued. Failure notification means for instructing switching, transaction business switching means for recovering the transaction business executed in the active system to a state in which it is possible to continue execution after receiving a transaction business switching instruction from the failure notification means, and after transaction business is restored Subsequent Transa Transaction business alternative processing means that takes over and continues the operation of the transaction business, batch business processing means that executes the batch business during normal operation, and monitors the business load of other computers at regular monitoring time intervals. First work load information monitoring means for recording work work load information of a computer in the work work load information table, when execution of transaction work is switched due to a failure in the computer and transaction work is resumed by the transaction work alternative processing means And comparing the work load information of another computer with the work load information of another computer by referring to the work load information table, and when it is detected that the work load of the other computer has increased, another Determining whether to switch batch jobs to a computer Switch job switching determination means, batch job switch notification means for instructing the other computer to switch the batch job when it is determined to switch the batch job to the other computer, and the work load of the computer for a certain monitoring time Second work load information monitoring means for monitoring the work load information of the computer at intervals and recording the work load information of the computer in the work load information table, batch work switching processing for switching the batch work instructed to be switched by another computer from the other computer And a program for operating the batch job switched by the batch job switching processing means as a batch job alternative processing means for continuing execution after the handover. 12. A transaction business processing means for executing transaction business as an active system during normal business operation of a computer, and whether or not the transaction business can be continued in the active system when a failure occurs in the computer. Failure event determining means for determining whether the transaction operation switching means existing in another computer is notified of the failure when the failure event determining means determines that the transaction operation in the active system cannot be continued. Failure notification means for instructing switching, transaction business switching means for recovering the transaction business executed in the active system to a state in which it is possible to continue execution after receiving a transaction business switching instruction from the failure notification means, and after transaction business is restored Subsequent Transa Transaction business alternative processing means that takes over and continues the operation of the transaction business, batch business processing means that executes the batch business during normal operation, and monitors the business load of other computers at regular monitoring time intervals. First work load information monitoring means for recording work work load information of a computer in the work work load information table, when execution of transaction work is switched due to a failure in the computer and transaction work is resumed by the transaction work alternative processing means To compare the work load information of another computer with the work load information of another computer by referring to the work load information table,
When it is detected that the business load of the other computer is high, another computer with a low business load is identified, and it is determined whether to switch the batch work to the other computer. Batch job switching Judgment means, batch operation switch notification means for instructing the other computer to switch the batch job when it is determined to switch the batch job to the other computer, and monitoring the work load of the computer at regular monitoring time intervals Second work load information monitoring means for recording work load information of the computer in the work load information table, batch work switching processing means for switching the batch work instructed to be switched by another computer from the other computer, and The jobs that were switched by the batch job switching processing means Program for operating as a batch job substitute processing means for continuing the execution of a subsequent take over switch operations.