JPS62159245A - Health checking system for computer system - Google Patents

Abstract

PURPOSE:To detect early a no-communication state by deciding a trouble occurring at the reception side in such a case where the transmission side transmits repetitively the start signals at a fixed time interval and in a fixed frequency after a data link is broken and then no answer is received from the reception side. CONSTITUTION:Both processors 1 and 2 are connected with each other by a public circuit containing an exchange network 5 via communication controllers 3 and 4. The processor 1 sends a start signal to the processor 2 and the processor 2 sends an answer signal back to the processor 1. An end signal is transmitted from the processor 2 after the end of a data link and the connection of the public circuit is cut off. The processor 1 transmits new start signals to the processor 2 at a fixed interval and in a fixed frequency in case the processor 1 receives no communication end signal nor the start signal of the next signal within a fixed period of time after the end of the data link. Then the processor 1 decides a trouble of the processor 2 and cuts off the connection of the public circuit if the answer signal is not received yet. In such a way the useless connection of the public circuit is prevented, therefore, the using cost of the public circuit is saved.

Description

[Detailed Description of the Invention] [Field of Application of the Invention] The present invention relates to a health check of the state in which data can be sent and received between computer systems connected via a public line,
In particular, between data transmission and reception (breaks in data links)
This paper relates to a health check method for computer systems.

[Background of the invention]

The conventional health check method for computer systems is
As described in Japanese Patent Application Laid-Open No. 54-105937, when communication is carried out between two or more computer systems using a public line, a state of no communication within the data link is monitored.

However, in order to ensure smooth operation between systems, conventional health checks of computer systems are performed only from the perspective of monitoring the operating status during data transmission and reception (within data links) between systems to determine whether they are normal or abnormal. Therefore, the no-communication state was not monitored after the data link was terminated. In other words, after one data link ends, one public @ line remains connected, and the receiving system waits for the next data operation request or termination signal to be transmitted from the transmitting system. At this time, if the sending system fails and there is no communication, the receiving system will continue to wait.9 In this way, when the sending system fails.

In addition, public lines were left unnecessarily connected, and no consideration was given to saving costs on public lines.

[Purpose of the invention]

SUMMARY OF THE INVENTION An object of the present invention is to provide a computer system health check method that can detect a no-communication state at a break in a data link between computer systems connected via a public line.

[Summary of the invention]

In communication between computer systems connected via a public line, the present invention provides that when a transmitting system transmits an activation signal and a receiving system receives the activation signal,
After the data link is disconnected because the receiving system sends a response signal using the standard communication procedure, the transmitting system sends the activation signal a certain number of times at fixed time intervals.
This is a health check method for computer systems that is characterized in that if there is no response from the receiving system after tl# is returned, it is determined that the receiving system is malfunctioning and the connection to the public line is disconnected.

[Embodiments of the invention]

An embodiment of the present invention is shown in FIGS. 1 and 2 below.

This will be explained with reference to FIG.

FIG. 1 shows a schematic diagram of computer systems connected to each other via a public line.

In the automatic system, a processing device 1 is connected to a partner system via a communication control device 3 over a public line having a switched network 5. Similarly, in the other system, the processing device 2
is connected to its own system via a communication control device 4 via a public line.

Next, regarding the signal transmission/reception status and health check method in this system, Fig. 2 shows the transmission and reception of signals from the own system and the other system, and the program for health check of the computer system performed in the communication control devices 3 and 4. The flowchart is shown in FIG. 3 and explained below.

This system is either one of the system is activated signal 2
This utilizes a communication procedure (SYN synchronous communication procedure) in which a response signal is sent back when the other system receives the activation signal.

To start communication between the local system and the other system, the local system dials and receives a start signal (
ENQO) is issued, and when the other system sends an activation signal (ENQO), the own system receives the activation signal (ENQ o), and in response, sends a response signal (A
CKo). The other system sends a response signal (AC
After receiving KO), 5IG, which is a notice of data communication.
Send a NON statement. The local system waits for the 5IGNON statement from the other system, and sends the 5IGNON statement.
When N sentences are received, a response signal (ACKl) is further transmitted to the partner system. The partner system that receives the response signal (ACKl) sends the data. When the transmission of this data is completed, the own system transmits a response signal (ACK2). Upon receiving the response signal (ACK2), the partner system transmits an end signal (EOTO) to its own system.

Then, the own system receives the end signal (EOTO) and confirms the end of data transmission. This completes the data link.

After the end of the data link, if there is no data to send, 5IG indicates that the other system wants to end the communication.
It sends a NOFF statement.

When the own system receives this 5IGNOFF statement, it will disconnect the public line from the other system.

On the other hand, if there is data to be transmitted after the data link ends, the other system sends a start signal for the next data communication.When the own system receives the start signal, Sends a response signal to the other party's system. When the partner system receives the response signal, it will start data transmission, and when the data transmission is finished,
The second data link will be terminated.

Since the present invention detects a non-communication state at a break in a data link, a health check is started after the data link ends (step 100).

After the data link ends, a health check starts (
100), if the other system is normal, it will either send a 5IGNOFF statement to end the communication to its own system, or it will send an activation signal @ for the next data transmission, as described above. (i, Jl) O5 IGNOFF statement tear (111), and since the communication has ended, processing is performed to disconnect the public line 1f (110), and the health check program ends (112). 5IGN
If it is an activation signal instead of an OFF statement (111), the health check program ends (112), and the next data communication is performed.

However, if the above-mentioned normal communication is not performed after the data link ends, the following processing is performed.

Own system within 3 seconds after the data link ends.

Activation signal (ENQl) or 5I from the other system
When no GNOFF statement is received (101
) initializes the transmission counter (C). In other words, C=0 (102). Next, in order to determine whether the other system is normal or abnormal, the own system transmits an activation signal (ENQ2) to the other system (103).

If the own system receives a response signal (ACK3) from the partner system within 6 seconds (104), it sends an end signal (EOTl)'f: to the partner system (108). A response in response to this end signal (EOTl).

When the response signal is sent from the partner system and the own system receives the response signal (109), it performs the process of disconnecting the public line for communication termination 4 (110) and ends the health check ( 112). Also, the own system sends an activation signal (ENQ 2) to the other system (103), and within 3 seconds a response signal (3 to Ac') is sent to the other system.
Even if the end signal (EO) is received (104), the end signal (EO
Tl) (108), and if no response signal is received (109), the process continues to the process for no response (105).

In addition, the local system sends an activation signal (ENQ2) to the other system (103), and within 3 seconds, a response signal (AC
K3) is not received (104), processing for no response is performed. Specifically, we measure how many times there is no response, and if there are more than 7 times there is no response (105).
, determine that the other party's system has failed and disconnect the public line (107), the health check program ends (112),) L, if no response occurs within 7 times (105) , 1 is added to the transmission counter (C), that is, C=C+1 (100). Then, the local system again sends the activation signal (ENQ4 or E
NQ5) to the partner system 103), a 3 second response signal (ACK4 or ACK5)'! ) waiting (104
), processing for non-response (ios) is performed.

Although the present invention relates to health checks between computers, it can also be applied to processing devices and peripheral devices within the same system.

According to this embodiment, a failure in the other system is detected early and the connection to the public line is disconnected, resulting in savings in line usage charges and further benefits such as being able to transfer the line to another system. There is.

〔Effect of the invention〕

According to the present invention, there is an effect that a non-communication state can be detected at an early stage at a break in the data link between computer systems connected through a public line.

[Brief explanation of drawings]

Figure 1 is a schematic diagram of the Funvita systems connected via a public line, Figure 2 is a diagram showing the transmission/reception status and health check method B between the own system and the other system in a computer system connected via a public line. FIG. 3 is a flowchart of a health check in a computer system connected via a public line. 1...Processing device of own system 2...Processing device of partner system 3...Communication control device of own system 4...Communication control device of partner system 5...Effective exchange network ENQ...Start Signal ACK...response signal

Claims (1)

[Claims] 1. When one sending computer system sends an activation signal to another receiving computer system between a plurality of computer systems connected via a public line, the receiving computer system, upon receiving the activation signal, A health check is performed during the data link by taking advantage of the convention that a response signal is returned to the sending computer system, and after the data link ends, the receiving computer system returns a response signal to the sending computer system. A communication end signal is sent to
In the communication control method for disconnecting the public line,
After the data link ends, the sending computer system waits for the receiving computer system to send a communication end signal or a start signal for starting the next data transmission within a certain period of time, Alternatively, if the activation signal is not received, the sending computer system sends a new activation signal to the receiving computer system a certain number of times at regular intervals, and if there is still no response, the sending computer system disconnects the public line connection. A computer system health check method characterized by disconnection.