JPH09305416A - Queue management system for inter-process communication - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a queue management system which can detect abnormality of transfer of queue information between processes in its early stage. SOLUTION: A queue management library 5 registers queue resource integral values (x) by receiving processes 3 and 4 and queue resource integral values (y) of all processes 1-4 in a queue database 6, outputs a message of process abnormality when an integral value (x) exceeds an abnormality prediction queue resource number α, and forcibly ends and restarts the process when the integral value (x) exceeds an abnormality limit queue resource number β and outputs a message of queue information on the process. When an integral value (y) exceeds an abnormality prediction queue resource number γof all the processes, system abnormality is alarmed with a message and queue transmission to a dummy process is performed; and queue information is recorded and a failure in transmission is sent back to transmitting processes 1 and 2.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a queue management system for interprocess communication in a supervisory control system or the like.

[0002]

2. Description of the Related Art Generally, a queue is used for inter-process communication, and a process number of a transmitting side, a process number of a receiving side process, transmission data, etc. are set in a predetermined format as queue information, and a process is exchanged between processes. I do.

At this time, the queue management information is also set at the same time, and the queue management library for controlling the queue between the processes refers to this to synchronize the processes to smoothly operate the system.

[0004]

Since the transfer of queues used for interprocess communication consumes system resources, the number of queues is limited. Normally, the number of queues that can be used for each system is limited. It is set.

Further, the queue controls the priority of processing in the queue management library by the queue management information included in the queue.

However, if queue transmission to a process that cannot receive a queue continues due to some problem, when resources used for interprocess communication are used up, queue transmission cannot be performed and operation of the entire system is stopped. become.

Further, even if resources are not exhausted,
If the priority queue is not processed due to a problem in the sending process, the process waiting for a response will also be in a waiting state for processing and will not be able to function as a system.

An object of the present invention is to provide a queue management system capable of early detecting an abnormality in the transfer of queue information between processes.

Another object of the present invention is to provide a queue management method capable of detecting a normal / abnormal state of a process and avoiding a system down by detecting an abnormal transmission / reception of queue information between processes.

[0010]

According to the present invention, in managing inter-process communication with a queue management library, the queue management library has a queue resource integrated value x for each receiving process and a queue resource integrated value y for all processes. Is registered in the queue database, and when the integrated value x is smaller than the abnormal limit queue resource number β for each receiving process and exceeds the abnormally predicted queue resource number α, a process error is output as a message, and the integrated value x is When the number of queue resources β is exceeded, the process is forcibly terminated and restarted, the queue information of the process is recorded and a message is output, and the integrated value y exceeds the error prediction queue resource number γ of all processes. Alerts you to a system error with a message, sends a queue to a dummy process for processing, and Record the transmission process is characterized in that the return transmission failure in the transmission process when the response waiting.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows the configuration of a queue management system showing an embodiment of the present invention. Of the processes 1 to 4 in the system, processes 1 and 2 are queue information transmission side processes, and processes 3 and 4 are queue information reception side processes.

The queue management library 5 manages resources and the like in the communication of queue information from the process on the transmission side to the process on the reception side, and registers this management information in the queue database 6.

The queue management information has a data structure having values shown in the table below.

[0014]

[Table 1]

In this table, α, β, and γ are set values determined from the importance of the process, the frequency of queue occurrence, and the like.

In the above configuration, the queue management library 5 predicts a system abnormality by quantitatively analyzing queue resource information for communication management between processes,
Furthermore, the system down is avoided from the abnormality prediction. The processing for abnormality prediction and system down avoidance will be described below.

(1) As shown in FIG. 2, the queue management library 5 determines that the queue resource integrated value x is smaller than the abnormal limit queue resource number β (S1) and is equal to or larger than the abnormality prediction queue resource number α (S2). , A warning is issued by a message or the like regarding the process abnormality (S3).

Therefore, when the integrated value x of queue resources of a certain receiving process approaches the abnormal limit queue resource number β, that is, the queue for a specific process is abnormally accumulated, or a queue resource longer than the time required for normal processing is used. If it is not released, a message can be issued to warn of an abnormal transfer of queue information.

(2) As shown in FIG. 2, when the queue resource integrated value x becomes equal to or larger than the abnormal limit queue resource number β (S1), the queue management library 5 forcibly terminates and restarts the process (S4). , S6). At this time, queue information of the equal process is recorded and a message is output (S5).

Therefore, when the queue resource integrated value x of a receiving process exceeds the abnormal limit queue resource number β, the queue resource can be released by forcibly ending the process.

In some cases, restarting the process may restore the system to a normal state and prevent the system from going down. At this time, the queue information can be returned to the queue transmission state before the process is restarted, and the cause of the abnormality can be analyzed from the stored information in the queue database.

Further, by parallel processing with the above processing (1), the queue resource integrated value x can be processed with a stepwise condition by the numbers α and β, and the stepwise processing enables abnormality prediction, In some cases, a hardware abnormality can be detected before the system goes down.

(3) As shown in FIG. 3, when the queue resource integrated value y in all processes is greater than or equal to the abnormality prediction queue resource number γ in all processes (S11), the queue management library 5 sends a message indicating a system abnormality. With a warning (S1
2) Then, the low priority queue is transmitted to the dummy process to be processed, the queue resource is released (S13), and the queue information is recorded (S14).

Thereafter, it is checked whether or not the sending process is waiting for a response (S15), and if it is waiting for a response, a queue transmission failure is returned and the queue resource is released (S16).

Therefore, as for the queue resource integrated value y for all processes of the system, similarly to the above (1) and (2), when the abnormality prediction queue resource number γ is exceeded, a message indicating that the queue information is exchanged is output. By doing so, a process abnormality can be warned.

Also, the queue resource can be released by transmitting to the dummy process. Furthermore, by returning the transmission failure to the transmission process, the queue resource is released and the system down can be avoided.

[0027]

As described above, according to the present invention, interprocess communication is performed while the queue management library manages the queue resource integrated value x for each receiving process and the queue resource integrated value y for all processes, and the receiving process is performed. Since an error in the number of queue resources in each process and all processes is detected and a message is output, the process is terminated forcibly, and the process is sent to a dummy process, the queue resource is secured by releasing the queue resource and the process error Communication that avoids detection and system down is possible.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a queue management system showing an embodiment of the present invention.

FIG. 2 is a queue management flow (1) according to the embodiment.

FIG. 3 is a queue management flow (2) according to the embodiment.

[Explanation of symbols]

 1, 2 ... Transmission process 3, 4 ... Reception process 5 ... Queue management library 6 ... Queue database

Claims (1)

[Claims] 1. When managing inter-process communication with a queue management library, the queue management library registers a queue resource integrated value x for each receiving process and a queue resource integrated value y for all processes in a queue database, When the integrated value x is smaller than the abnormal limit queue resource number β for each receiving process and exceeds the abnormal predictive queue resource number α, a process error is output as a message, and the integrated value x exceeds the queue resource number β. Then, the process is forcibly terminated and restarted, the queue information of the process is recorded, and a message is output. When the integrated value y exceeds the error prediction queue resource number γ for all processes, a system error is displayed as a message. Warn, send a queue to the dummy process to process it, record queue information, send process A queue management method for inter-process communication characterized by returning a transmission failure to the transmission process while waiting for a response.