JPH06290126A - Fault monitoring system for computer system - Google Patents

Abstract

PURPOSE:To early and surely detect the fault of a client machine and to recovery the fault by providing a monitoring process on the client machine to monitor the state of a client process. CONSTITUTION:By providing a client machine with a monitoring process, and a server machine with a monitoring process check request means 12, respectively, the state of the client machine is monitored. The means 12 periodically transmits the requests to the processing process of the client machine for notification of the client machine state independently of the general work processing. The monitoring process of the client machine receives the state notification request from the server machine and checks the state of the client process to notify the server machine side of this checking result. If the answer contents includes the fault occurred in the client process or if no answer is received, the occurrence of a fault is decided and the recovery processing is carried out for the fault.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a failure detection method in a network system, and more specifically, in a client / server system composed of a server machine and client machines connected to each other, a client The present invention relates to a failure detection method for detecting and recovering a failure that has occurred on a machine and maintaining consistency as a system and continuity of data processing.

[0002]

2. Description of the Related Art Conventionally, in a client / server system, a client program is composed of one process. Therefore, the failure detection by the server machine is performed when the failure report is received from the client process while the server machine is waiting for the request from the client process, or the processing result is returned from the server machine to the client machine. When a failure of the client machine including the network transmission path is detected at the time of, there is a limit to the timing at which the failure can be detected.

A fault detection method in the conventional system will be described below with reference to FIG. The process in the server process waits for a processing request from the client process (step 1), branches to divide the process depending on whether the message from the client is a processing request or an abnormal state (step 2), and responds to the processing request from the client. Execution of processing (step 3), transmission of the processing result by the server process to the client process (step 4), branching depending on whether the transmission was performed normally (step 5), recovery processing when an abnormality is detected (step 6) ). On the other hand, the processing in the client process is in a state of waiting for a processing request from the user (step 7), receiving an input from the user (step 8), processing request to the server process (step 9), and waiting for a processing result from the server process. (Step 10),
The process is performed by the client process (step 11) on the processing result from the server process. In this manner, the server process waits for a processing request from the client process, when an abnormal state of the client machine is notified, or when the processing result is transferred from the server machine side, the server process Abnormality processing was performed only when abnormalities in the process were detected.

[0004]

In the client / server system configured as described above, there is a restriction on the timing of failure detection, and a failure occurs in the network while the server process is waiting for a processing request from the client process. In some cases, or in situations where the client process is unable to communicate the failure to the server process, the server process cannot detect the failure, and therefore may see this even if the resource remains locked. There was a problem that the system could not be recovered and the processing as a system lacked consistency and reliability.

The present invention has been made in order to solve the above problems, and a reliable client / server by detecting a failure occurring in a client machine reliably and early and executing recovery processing. The purpose is to provide a system.

[0006]

A failure detection processing method in a client / server system according to the present invention comprises:
A monitoring process for monitoring the status of the client process is provided on the client machine, and a check request (polling) means to the monitoring process for periodically requesting the client status information for the monitoring process is provided on the server machine. It was done. Further, the failure detection processing method in the client / server system according to the present invention is provided with means for periodically notifying the server machine of the status information of the client process on the client machine, and the client machine from the client machine is provided on the server machine side. A timer monitoring means for checking the lack of notification information regarding the process state is provided.

[0007]

In the fault detection system according to the present invention, the state of the server machine is monitored by the client machine on a regular basis, so that a fault occurring in the client machine can be detected early without overlooking it. You can Further, in the failure detection method according to the present invention, the client machine periodically notifies the server machine of the status information, and on the other hand, the server machine performs timer monitoring to detect the missing status information. With this, it is possible to detect a failure occurring in the client machine at an early stage.

[0008]

【Example】

Example 1. Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a processing outline of a failure detection processing method for monitoring the status of a client machine by providing a monitoring process on the client machine side and providing a monitoring process check request means on the server machine side. In the figure, 1 to 11 are the same as the corresponding numbers in the conventional method. Reference numeral 12 is a monitoring process check requesting unit that is activated at regular intervals to prompt the client machine to notify the client machine of the process status.
3 is a step of waiting for a status notification request from the monitoring process check requesting means, 14 is a step for checking the status of the client process and notifying the server machine, and the monitoring process on the client machine is It is composed of two steps 13 and 14.

Next, the operation will be described. On the client machine, when the client process is started, the monitoring process is also started at the same time. Between the client machine and the server machine, general business processing is executed by the client process (steps 7 to 11) and the server process (steps 1 to 6). Monitoring process check means 12 on the server machine
Independently issues a status notification request to the monitoring process on the client machine side, independently of the above general business processing. The monitoring process on the client machine also operates independently of the client process, receives the above status notification request from the server machine, checks the status on the client process side, and notifies the result to the server machine side. . If there is a failure in the client process in the response contents to the status notification check request from the server machine for the monitoring process on the client machine, or if there is no response to the status notification request, the server machine On the side, an abnormality recovery process is performed assuming that a failure has occurred in the client machine.

Embodiment 2. A second embodiment of the present invention will be described with reference to the drawings. In FIG. 2, means for periodically notifying the server process of the status information of the client process is provided on the client machine, and means for monitoring the notification of the status information from the client machine is provided on the server machine by a timer. It is a figure showing the processing outline of the failure detection system which monitors the state of a client process. In the figure, 1 to 11 are the same as the corresponding numbers in the conventional method. Reference numeral 15 is a client state notifying unit which is activated periodically to notify the client machine of the client process state, and 16 is for checking whether the client state information is periodically reported from the client state notifying unit. It is a timer monitoring means.

Next, the operation will be described. On the client machine, when the client process is started, the client status notification process is also started. Between the client machine and the server machine, general business processes are client processes (7-
11) and the server process (steps 1 to 6). The client status notification process 15 periodically notifies the server machine of the status of the client machine, independently of the above general business processing. On the server machine, the timer monitoring check means 16 checks whether or not the client status information is periodically sent. If the notification content reports that a client process failure has occurred, or a client process that has been delayed for notification or has not been notified for a certain period of time, considers that a failure has occurred and recovers from the failure. Perform processing.

[0012]

Since the present invention is constructed as described above, it has the following effects. The client process status can be changed without depending on the execution status of the general business process on the client machine.
It is possible to notify the server machine easily without delay. Also, by providing the timer check monitoring means on the server machine, the status information sent to the server machine in response to a status notification request from the server machine or periodically started on the client machine can be used. Detects even if there is a loss, and detects even intermittent failures on the client process side without failing to detect them early so that the continuity of data processing and the overall network system The reliability of can be increased.

[Brief description of drawings]

FIG. 1 is a diagram showing a processing outline of an abnormality processing method using a monitoring process.

FIG. 2 is a diagram showing a processing outline of an abnormality processing method using periodic notification of status information from a client side.

FIG. 3 is a diagram showing processing contents of a conventional method.

[Explanation of symbols]

 12 Check Request Means to Monitor Process 13 Status Notification Request Waiting Means 14 Client Process Status Check and Notification Means 15 Client Process Status Check and Notification Means 16 Timer Monitoring Means

Claims (2)

[Claims] 1. A fault monitoring system in a network-connected computer system, wherein the computer system is provided with a first computer on which a client business process is installed, resource management and service provision in the network system. And a client process status monitoring unit for monitoring and notifying the client process status operating on the first computer.
The computer system failure monitoring method, wherein the computer is provided with polling means for periodically requesting the client process status information to the client process status monitoring means. 2. A fault monitoring system in a computer system connected to a network, the computer system including a first computer having a client business process installed therein, resource management and service provision in the network system. The second computer is configured to monitor the client process status operating on the first computer, and periodically transfers the client process status information to the second computer. And a means for monitoring the client process state information sent from the first computer, which is regularly activated on the second computer. Computer system fault monitoring method.