JPH05250195A - Health check control system of information processing system - Google Patents

Abstract

PURPOSE:To prevent an erroneous recognition that a system down occurs by monitoring an opposite side processor by mutual data communi-cation and monitoring the processor by the file information of the magnetic disk device between processors. CONSTITUTION:Between a processor 11 working as an active system processor and a processor 12 working as a standby system, the monitoring of the mutual working state is performed by a notification telegraphic message 33 to be transmitted by a fixed time interval by a cyclic timer 32 from the processor 12 to the processor 11 and a response telegraphic message 34 to be transmitted from the processor 11 to the processor 12 for this notification telegraphic message 33. When the active system processor 11 is not capable of receiving due to the occurrence of a fault, the standby system processor 12 reads the file information stored in a magnetic disk device plural number of times and compares them. When the down of the processor 11 is detected by this comparison, the processor 12 is made to work as the active system by switching the processor.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a health check control system for an information processing system which employs a hot standby system having a duplex configuration, and particularly to a health check control for an information processing system which is operated remotely by unmanned operation. Regarding the scheme.

[0002]

2. Description of the Related Art A conventional health check control method for an information processing system that employs a hot standby method with a duplex configuration allows the other side to communicate only by mutual data communication via a communication line installed between two processors. Monitoring the operating status of the processor.

[0003]

As described above, the health check control system of the conventional information processing system which employs the hot standby system having the duplex configuration uses the communication line installed between the two processors. Since the operating status of the processor on the other side is monitored only by mutual data communication, it is impossible to send data from the active processor due to a failure of the communication control unit or communication line of the active processor. In this case, the standby system processor has a disadvantage that the active system processor mistakenly recognizes that the system is down.

Further, when the two processors are both started up as standby processors at the initial startup of the system, each of them operates normally and the other processors can recognize the normal operating status of each other. In this case, the predetermined processor performs the switching operation so that it operates as the active processor, and then operates as the active processor.However, due to some reason, data from the other processor can be received. If not, both of the two processors also have a drawback that they switch to the active processor.

[0005]

A health check control method for an information processing system according to the present invention is a health check control method for an information processing system that employs a hot standby method with a duplex configuration. Is monitored by the active processor by data communication via a plurality of communication lines installed between the active processor and the standby processor, and the system operator is notified when the standby processor is down. To that effect, the operating status of the active processor is communicated via a plurality of communication lines installed between the active processor and the standby processor, and the standby processor and the File information stored in the magnetic disk unit installed between the active processor and Wherein when detecting the down of the active system processor includes be operated as a processor of the active system by switching processor of the standby system by reading several times Te compare them.

[0006]

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

FIG. 1 is a schematic diagram showing an embodiment of the present invention, FIG. 2 is a block diagram showing an example of an information processing system to which the embodiment of FIG. 1 is applied, and FIG. 3 is a magnetic disk device of the example of FIG. 5 is a record format diagram showing an example of the contents of the file of FIG.

In the information processing system of FIG. 2, two direct connection lines 15a and 15b are connected by a female-female twist connector cable between the two processors 11 and 12 having a duplex configuration, and the primary and secondary circuits are provided between them. Two
The two magnetic disk devices 13 and 14 are connected.
The direct connection line and the magnetic disk device are two lines and two units, respectively, as a countermeasure against failure.

The files of the magnetic disk devices 13 and 14 are relative organization files having one record of 256 bytes, and in the record, as shown in FIG.
It includes primary / secondary type fixed information 21 indicating the primary / secondary type of the magnetic disk device, a counter value 22 updated by the active processor, and processor type information 23 indicating the type of the processor that has output the current value information. .. The active processor updates the file information of both the primary and secondary magnetic disk devices 13 and 14 at regular time intervals, and the standby processor 12 updates the information of the active system in the data communication via the direct connection line 15a or 15b. When the response is not sent from the processor 11, the information of the file in the positive magnetic disk device 13 is read and it is judged whether or not the file information read from the magnetic disk device 13 last time is updated. At this time, if the file information cannot be read because the primary magnetic disk device 13 has a failure, the file information of the secondary magnetic disk device 14 is read.

In this embodiment, as shown in FIG. 1, between the processor 11 operating as the active processor and the processor 12 operating as the standby processor, the processor 12 changes from the processor 12 to the processor 11. On the other hand, a notification message 33 transmitted at a constant time interval by the cyclic timer 32 and a response message 34 transmitted from the processor 11 to the processor 12 in response to the notification message 33 mutually monitor the operating state. It is carried out. Further, the active processor 11 updates the counter value 22 of the file information of the magnetic disk devices 13 and 14 at a fixed time interval by the cyclic timer 31, asynchronously with this. The timer value of the cyclic timer 31 of the active processor 11 is set to be smaller than twice the timer value of the cyclic timer 32 of the standby processor 12.

When a failure occurs and the active processor 11 cannot receive the notification message 33 from the standby processor 12, the response message 34 cannot be transmitted.
The standby system processor 12 cannot grasp the operating state of the active system processor 11. Therefore, if the processor 12 of the standby system cannot receive the response message 34 from the processor 11 of the active system even after the elapse of a certain time after the transmission of the notification message 33, the information of the file in the positive magnetic disk device 13 is obtained. Read again and notify again by another line 33
To send. If the response message 34 cannot be received after a certain period of time, the line is changed and the notification message 33 is transmitted again. Response message 3
4 cannot be received, the information of the file in the magnetic disk device 13 is read again and the previous magnetic disk device 1 is read.
The counter value 22 is compared with the file information read from 3. Since the timer value of the cyclic timer 31 of the active processor 11 is set to be smaller than twice the timer value of the cyclic timer 32 of the standby processor 12, if the active processor 11 is operating. , The standby processor 12 is the second time the magnetic disk device 13
The counter value 22 with the file information read from the magnetic disk device 13 should have been updated at least once from the counter value 22 with the file information read from the magnetic disk device 13 at the first time. Therefore, as a result of the comparison of the counter values 22 in the standby system processor 12, the active system processor 11
When it is determined that the counter value 22 has not been updated by, the standby-system processor 12 recognizes that the active-system processor 11 is down, and switches the standby system to the active system. After that, it operates as an active processor.

[0012]

As described above, according to the health check control method of the information processing system which employs the hot standby method with the duplex configuration of the present invention, the operating status of the processor on the other side is controlled by the mutual data communication via the communication line. In addition to the method for monitoring the standby system, the file information is provided in the magnetic disk device connected between them, and the counter value is updated at regular time intervals, so that the active processor of the active system is activated. By making it possible to investigate by using the standby system, even if the active processor is operating normally, if the data transmission from the active processor becomes impossible due to a communication line failure, etc. Can prevent the active processor from erroneously recognizing that the active processor is down. There is. Further, at the time of initial startup of the system, when data cannot be received from the processor on the other side for some reason, it is possible to prevent the two processors from both switching to the active processor. There is also.

[Brief description of drawings]

FIG. 1 is a schematic view showing an embodiment of the present invention.

FIG. 2 is a block diagram showing an example of an information processing system to which the embodiment of FIG. 1 is applied.

3 is a record format diagram showing an example of the contents of a file in the magnetic disk device of the example of FIG.

[Explanation of symbols]

 11 processor 12 processor 13 magnetic disk device 14 magnetic disk device 15a / 15b direct connection line 21 primary / secondary type fixed information 22 counter value 23 processor type information 31 cyclic timer 32 cyclic timer 33 notification message 34 response message

Claims (2)

[Claims] 1. An operating state of a standby system processor is monitored by the active system processor by data communication via a plurality of communication lines installed between the active system processor and the standby system processor, and the standby system is monitored. When a down state of the processor is detected, the system operator is notified of that fact, and the operating status of the active processor is set between the active processor and the standby processor. Data communication through the standby system and the standby system processor when the down of the active system processor is detected by monitoring by file information stored in the magnetic disk device provided between the standby system processor and the active system processor. Checking the health of an information processing system, including switching between the two to operate as the active processor control method. 2. In a health check control method for an information processing system that employs a hot standby method with a duplex configuration, the operating status of a standby system processor is set between an active system processor and the standby system processor. When the active processor monitors the active processor by data communication via a plurality of communication lines and detects that the standby processor is down, the system operator is notified of the fact and the operating status of the active processor is checked. Data communication via a plurality of communication lines installed between the active processor and the standby processor, and a file stored in a magnetic disk device provided between the standby processor and the active processor By reading the information a plurality of times and comparing them, the data of the active processor is reduced. A health check control method for an information processing system, which comprises switching the standby system processor to operate as an active system processor when an error is detected.