JPH1185714A - Execution resource control program for computer duplex system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the availability of a computer duplex system by communicating the execution/stop control of resources or the switching control of a computer to execute the resources with an execution resource control program. SOLUTION: When it is requested to switch the execution of all the resources under execution at a computer 20 to a computer 30, while communicating with a working state control program 214, an execution resource control program 211 starts processing for switching the computer to execute all the resources. When there is no resource under execution at the computer 20, a switching end notice is sent to the working state program 214. As a result, at the time point when the working state control program 214 stops working through a working managing interface, the execution of resources under execution at the computer 20 is completely switched to the computer 30 and since there is no resource under execution at the computer 20, even while an execution resource control program 311 detects non-working of the computer 20, the availability of resources presented by the computer duplex system is not lowered.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a dual computer system which provides high availability by using two computers.

[0002]

2. Description of the Related Art A conventional computer duplication system will be described with reference to the drawings.

FIG. 2 shows a conventional computer duplication system.

The computers 20 and 30 are computers constituting a redundant computer system. Since the computers have the same configuration in the redundant computer system, only the configuration of the computer 20 will be described. In the computer 20, an operating system 21 is executed.
The execution resource control program 211 has been executed above.
The computer 20 is equipped with communication means 23 and 24 for performing communication. The communication unit 23 can communicate with a computer connected to the network 10, and the communication unit 24 can communicate with a computer connected to the network 50. Further, a SCSI controller 22 for accessing a hard disk outside the computer is mounted.

The computer 60 is a management computer for the computer redundancy system. The computer redundancy system management software 611 is executed on the operating system 61. The computer 60 manages the computer redundancy system connected via the network 10 by the communication means 62. it can.

The computers 20 and 30 share the hard disk 40 by the SCSI controllers 22 and 32, read resources from the shared resource 41 in the hard disk 40 by the execution resource control programs 211 and 311, and execute and stop the resources on both computers. It is possible to control. FIG.
Then, the resource 212 is executed by the computer 20 and the computer 30
The resource 312 is being executed.

[0007] The resources executed by each computer are provided to users connected to the network 10 using the communication means 23 and 33. In FIG. 2, the resource 212 is executed by the computer 20, the resource 312 is executed by the computer 30, and each resource is provided to the user.

[0008] The computers 20 and 30 are connected by a network 50 by communication means 24 and 34, and can monitor each other's operation status by performing communication. When the communication means 24 detects that the computer 30 has become inactive for some reason, the execution resource control program 211 automatically reads the resource 312 being executed by the computer 30 from the shared hard disk 40 into its own computer. It is executed by its own computer in place of 30 to maintain the availability of the resource 312. Even when the computer 20 becomes inactive, the resources 212 are executed by the computer 30 in the same procedure, and the availability is maintained.

Further, the administrator communicates with the execution resource control program of the computer 20 or the computer 30 by using the management software 611 to request the execution / stop control of the resource or the switching control of the execution computer of the resource. It is possible to

[0010]

In the conventional computer redundant system, the switching of the execution computer of the resource by the execution resource control program is executed when the execution resource control program of one computer detects that the other computer is not operating. You. The only means for detecting the non-operation of the other computer is that the execution resource control programs of both computers mutually monitor each other by communication, and the time required for the execution resource control program to detect the non-operation state of the other computer is limited. This is the key to maintaining the availability of a redundant computer system.

When one computer normally terminates its operation, it is necessary to switch the execution of the resource being executed by this computer to the other computer. However, until the execution resource control program of the partner computer detects that this computer is not running,
The execution computer of the resource is not switched.

In this case, the resources that have been executed on the computer that has not been operated cannot be provided to the user until the computer is switched, so that the availability of the system is reduced.

Similarly, when the operation of one computer is stopped due to a serious hardware failure, it is necessary to switch the execution of the resource being executed by this computer to the other computer. Until the control program detects that the operation of this computer has stopped, the execution computer of the resource is not switched. For this reason, until the switching of the computers is performed, the resources that have been executed on the computers that have become inactive cannot be provided to the users, and the availability of the system is reduced.

Further, depending on the operation mode of the redundant computer system, there is a case where the execution of resources is stopped to reduce the load on the computer in order to increase the processing performance at a specific time.
In this case, the administrator must manually control the resources using the management software of the redundant computer system, which increases the burden on the administrator.

An object of the present invention is to solve the above problems, improve the availability of a redundant computer system, and automate execution resource management of the redundant computer system.

[0016]

The object of the present invention is to provide a means for externally requesting and executing control of resource execution / stop and switching of a computer executing a resource by communicating with an execution resource control program. A means for automatically executing the above means by an external program such as a computer operating state control program, a hardware failure monitoring program, and a means for the execution resource control program to automatically perform the above control by control execution schedule information. Achieved.

[0017]

Embodiments of the present invention will be described with reference to the drawings.

FIG. 1 shows a computer duplication system including two computers. The computers 20 and 30 are computers constituting a redundant computer system. Since the respective computers have the same configuration in the redundant computer system, the computers 20 and 30 have the same configuration.
Only the configuration of 0 will be described.

The computer 20 has a configuration similar to that of a conventional computer redundant system, and additionally, an operating state control program 214 and a hardware fault monitoring program 213 are executed on the operating system 21.

As shown in FIG. 3, the operating state control program 214 is communicated with the computer management software 612 running on the operating system 61 of the remote computer 60, and is requested by the administrator to terminate the operation of the computer. , The operation management interface 215
It has a function to end the operation of.

Hardware failure monitoring program 213
As shown in FIG. 4, by reading the hardware failure information written by the driver 26 from the hardware failure information 271 on the local hard disk 27, CPU temperature abnormality of the computer, memory degeneration, enclosure temperature rise, voltage drop It has a function to detect hardware failures such as.

The execution resource control program 211 shown in FIG.
As shown in FIG. 5, the management interface 216 of the redundant computer system has a function of controlling the execution / stop of the resources provided in the redundant computer system and controlling the switching of the executing computer of the resources. In addition, the execution resource control program 211
3. It has a communication means with the hardware failure monitoring program 214, and has a function of executing the control when each program requests the above control. Further, as shown in FIG. 5, the execution resource control program 211
8 has a function of referring to the schedule information 281 and automatically executing the control set in the schedule information.

A description will now be given of a computer switching method when one of the computers of the redundant computer system is normally terminated in response to a request from the computer management software of a remote computer. Here, the administrator uses the computer management software 612 running on the computer 60 to
Try to end the operation of.

The operating state control program 214 at this time
FIG. 6 shows the flow of the process of FIG. 6, and FIG. 8 shows the flow of the process of the execution resource control program 211.

First, the flow of processing of the operating state control program 214 will be described.

When the operation of the computer 20 is stopped, the operating state control program 214 executed on the computer 20
Is the computer management software 612 of the remote computer 60
Receives a computer operation end request from. Then 6010
Starts the computer operation end processing. The operation status control program 214 is executed by the operation management interface 215 before executing the computer.
Communication is performed with 11 and a request is made at 6020 to switch the execution computers of all the resources being executed by the computer 20 to the computer 30. After that, in 6030, the end of switching of the execution computer is received from the execution resource control program 211, and after confirming that the resource is not being executed by the own computer, 604
At 0, the operation of the computer 20 is terminated by the operation management interface 215.

Next, the processing flow of the execution resource control program 211 will be described.

When the execution resource control program 211 is requested to switch execution of all the resources being executed by the computer 20 to the computer 30 through communication with the operation status control program 214, the execution resource control program 211 Start processing. If there is a resource being executed by the computer 20 in 8020, the execution of the resource is switched from the computer 20 to the computer 30 in 8030. Calculator 2 at 8020
Until there are no resources being executed at 0, the switching process of the execution computer is continued. When no resources are being executed by the computer 20, a switch completion notification is sent to the operating state control program 214 in 8040.

As a result, when the operation state control program 214 of the computer 20 terminates the operation by the operation management interface, the execution of the resources executed by the computer 20 is completely switched to the computer 30 and the computer 20 is executing the resources. Therefore, the availability of resources provided by the redundant computer system does not decrease while the execution resource control program 311 of the computer 30 detects that the computer 20 is not operating.

Next, a description will be given of a computer switching method in the case where the housing temperature of one computer of the redundant computer system exceeds a threshold value. Here, it is assumed that the housing temperature of the computer 20 has exceeded the threshold value.

As shown in FIG. 4, the hardware failure monitoring program 213 reads the hardware failure information written by the driver 26 from the hardware failure information 271 on the local hard disk 27, thereby increasing the temperature of the computer housing. It has a function to detect hardware failures.

The hardware failure threshold information 272 on the local hard disk 27 is referred to by the driver 26, and when the chassis temperature exceeds the threshold, the hardware failure information 271 is added to the hardware failure threshold information. A hardware failure is written and the hardware failure monitoring program 213 is notified of the failure. The hardware failure threshold information 272 can be set by the hardware failure monitoring program 213.

Further, a hardware failure type can be set in the computer switching information 273 on the local hard disk 27, and the hardware failure monitoring program 2
13 indicates that the detected failure information is the computer switching information 273.
If it is set to, the computer switching request processing is executed. The computer switching information 273 can be set by the hardware failure monitoring program 213.

Here, the administrator communicates with the hardware failure monitoring program 213 of the computer 20 using the computer management software 612 of the remote computer 60, and the case temperature threshold of the hardware threshold information 272. It is assumed that the value is set to 40 degrees, and the hardware failure of the case temperature rise is set in the failure type of the computer switching information 273.

FIG. 7 shows the processing flow of the hardware failure monitoring program 213 when the housing temperature exceeds 40 degrees, and FIG. 7 shows the processing flow of the execution resource control program 211. 8

First, the hardware failure monitoring program 2
13 is described.

The local hard disk 2 of the computer 20
7 is 40 ° C., the hardware failure monitoring program 213 transmits the hardware failure threshold value information 271 on the local hard disk 27 via the hardware failure information 271 on the local hard disk 27. Then, a hardware failure notification that the housing temperature has exceeded 40 degrees is received from the driver 26. Thereafter, the hardware failure monitoring program 213 starts processing at the time of hardware failure occurrence at 7010. With reference to the computer switching information 273 on the local hard disk 27, it is determined at 7020 whether or not a hardware failure due to a rise in the enclosure temperature is set as a hardware failure type. In this case, since the hardware failure of the case temperature rise is set in the computer switching information 273, the computer communicates with the execution resource control program 211 in 7030, and executes the execution computers of all the resources being executed in the computer 20 in the computer 30. Request to switch to Thereafter, in 7040, the execution resource control program 211
, And confirms that the resources are not being executed by the computer 20. If 7020
If the hardware failure due to the rise in the enclosure temperature is not present in the computer switching information, the computer does not request the execution resource control program to switch the computer, and ends the hardware failure processing at 7050.

Next, the processing flow of the execution resource control program 211 will be described.

The execution resource control program 211 executed by the computer 20 is the hardware failure monitoring program 2
13 is requested to switch execution of all resources being executed by the computer 20 to the computer 30 by communication with the
At 8010, the all resource execution computer switching process starts.
At 8020, if there is a resource being executed on the computer 20,
At 8030, the execution of the resource is sent from the computer 20 to the computer 30.
Switch to Until the resources being executed by the computer 20 at 8020 are exhausted, the execution computer switching process is continued. When no resources are running on the computer 20, the hardware failure monitoring program 213 is executed at 8040.
Send a switch end notification to.

Thereafter, when the housing temperature of the computer 20 continues to rise and the operation of the computer 20 is stopped due to a hardware failure, the execution of the resources already executed by the computer 20 is switched to the computer 30. Since no resources are being executed by the computer 20, the availability of the resources of the redundant computer system does not decrease while the execution resource control program 311 of the computer 30 detects that the computer 20 is not operating.

Lastly, a description will be given of an automatic system for managing execution resources in a computer redundant system. Here, an automatic execution method of the execution resource control program 211 executed by the computer 20 will be described.

As shown in FIG. 5, the execution resource control program 211 reads control information from the schedule information 281 on the local hard disk 28 and automatically executes the execution resource control. The schedule information 281 can be set by the execution resource control program 211. Here, in order to improve the processing performance of the nighttime batch processing, the administrator sets up the computer duplication system management software 6 of the remote computer 60.
11 and the execution resource control program 2 of the computer 20
At 11:50 pm when the batch process is started, the resources being executed by the computer 20 are switched to the computer 30 at 10:50 pm in order to eliminate the resources being executed by the computer 20 at 11 pm when the batch processing is started. It is assumed that the current time is 10:50 pm.

The execution resource control program 211 at this time
FIG. 9 shows the flow of the processing of FIG.

The execution resource control program 211 starts a schedule information reference process at 9010 periodically.
The execution resource control program 211 refers to the schedule information 281 on the local disk 28 in 9020, and
At 030, it is determined whether the current time is set in the schedule information. Since the current time of 10:50 pm is set in the schedule information 281, the control information of the schedule information is referred to. In this case, the control information includes the computer 20
Since the execution of the resource is set to be switched to the computer 30, the execution computer of all the resources being executed by the computer 20 is switched from the computer 20 to the computer 30 according to the control information in 9040. If the current time is not set in the schedule information at 9030, the schedule information reference processing is terminated at 9050, and after a predetermined time, the schedule information reference processing is started again.

At 11:00 pm when batch processing is started by the automatic execution processing of the execution resource control program 211,
The execution of the resources already executed by the computer 20 is all switched to the computer 30, and there is no resource executed by the computer 20, so that the processing performance of batch processing is improved.

This concludes the description of the embodiment of the present invention.

[0047]

Conventionally, in a redundant computer system, computers are not switched until the normal operation end of one computer or the operation stoppage due to a hardware failure is detected by the execution resource control program of the other computer. The resources that were running on the computer that was not running could not be provided. However, by requesting the execution of the execution resource control program from an external program, it is possible to provide resources even while one of the computers is not operating, thereby improving the availability of the redundant computer system.

Further, by switching computers based on the schedule information set by the administrator, the operation of the redundant computer system is automated and the burden on the administrator is reduced.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a redundant computer system when an operating state control program and a hardware failure monitoring program cooperate with an execution resource control program.

FIG. 2 is a configuration diagram of a conventional computer redundant system.

FIG. 3 is a configuration diagram of an operating state control program and remote computer management software.

FIG. 4 is a configuration diagram of a hardware failure monitoring program and remote computer management software.

FIG. 5 is a configuration diagram of an execution resource management program when resource management is automated by schedule information.

FIG. 6 is a flowchart of processing of an operation state control program when a computer operation end request is received.

FIG. 7 is a flowchart of processing of a hardware failure monitoring program when a hardware failure occurs.

FIG. 8 is a flowchart of processing of an execution resource control program when an all resources execution computer switching request is received.

FIG. 9 is a flowchart of a schedule information reference process;

[Explanation of symbols]

20: Computer, 21: Operating system, 21
3: Hardware failure monitoring program, 214: Operating state control program.

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Hiroyuki Igata 5030 Totsukacho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Software Development Division, Hitachi, Ltd.

Claims (14)

[Claims] 1. A computer shared by two computers connected by a network executes resources shared by the two computers,
When one computer becomes inactive, another computer takes over execution of resources in a computer duplication system, which is an execution resource control program that controls switching of computers. A resource execution / stop control unit, a resource execution computer switching control unit, and communicates with an operation state control program and a hardware failure monitoring program that run on both computers by the management interface of An execution resource control program for executing control according to the following. 2. The operation state control program has a function of stopping operation of a computer by an operation management interface when an operation end is requested by communication with computer management software of a remote computer. Computer operating state control method. 3. The operation state control program according to claim 2, when receiving an operation end request from the computer management software of the remote computer, communicates with the execution resource control program to execute all the resources being executed by the own computer. A computer for requesting a switch to a partner computer, confirming that there is no resource being executed on the own computer by a notification from the execution resource control program, and then stopping the operation of the computer by the operation management interface. Operating state control method. 4. The hardware fault monitoring program according to claim 1,
A hardware failure monitoring method having a function of detecting a hardware failure by sharing a database mechanism in a local hard disk with a driver and receiving a hardware failure notification from the driver. 5. The hardware fault monitoring program according to claim 4, wherein a threshold value of a case temperature abnormality and a voltage abnormality notified by the driver can be set in a database in a local hard disk shared with the driver. A hardware fault monitoring method characterized by detecting that the housing temperature and voltage exceed set values as fault information. 6. The hardware failure monitoring program according to claim 4, wherein the hardware failure monitoring program detects a hardware failure such as a CPU temperature abnormality, a memory degeneration, a case temperature rise, a voltage drop, etc., based on a notification from the driver. Based on the switching information stored in the mechanism, it communicates with the execution resource control program, requests that the execution of all resources being executed on its own computer be switched to the partner computer, and before the computer stops operating due to a hardware failure. A computer switching method characterized by eliminating resources being executed by the computer. 7. The switching information according to claim 6 is stored in a database in a local hard disk, and information can be set by the computer redundancy system management software of the remote computer communicating with the hardware failure monitoring program. A computer switching method characterized in that the switching timing at the time of a hardware failure can be set freely according to the hardware environment. 8. The switching information according to claim 6, wherein a hardware failure type such as CPU temperature abnormality, memory degeneration, enclosure temperature rise, voltage drop and the like can be set. A computer switching method characterized by communicating with an execution resource control program and issuing a computer switching request when a set hardware failure is notified from a driver. 9. The switching information according to claim 6, wherein the number of fault occurrences can be set, and the hardware fault monitoring program is configured to execute the hardware fault setting set in the switching information when the set number of times has occurred. A computer switching method for communicating with an execution resource control program and issuing a computer switching request. 10. The execution resource control program according to claim 1,
An execution resource control method having a function of communicating with a remote computer redundant system management software and executing an execution resource control when a control request is received from the management software. 11. The execution resource control program according to claim 1,
Execution resource control characterized by periodically referring to schedule information stored on a hard disk, executing execution resource control of a redundant computer system based on the schedule information, and automating execution resource management of the redundant computer system. method. 12. The schedule execution function according to claim 11, wherein the schedule information has time information, and controls execution resources at a set time. 13. The schedule execution function according to claim 11, wherein execution / stop can be set for each resource, and control is performed to execute a specific resource and stop execution of another resource. . 14. The schedule execution function according to claim 11, wherein an execution computer can be set for each resource, and the distribution of resources executed by each computer is controlled.