JPH03262350A - Communication control system for highly reliable local area network - Google Patents

Abstract

PURPOSE:To transfer a message to an object sub system without a delay even when a fault takes place in any of transmission lines by transferring a same data to plural transmission lines almost simultaneously. CONSTITUTION:A transmission message identifier and a message form identifier are given to a data to be transferred to a sub system B by a mask system CPU A1 of a subsystem A. Then a message is sent almost simultaneously to a group of nodes B11-B13, B21-B23 of the subsystem B through transmission enable LAN transmission lines 1-3. Thus, the same data is sent to the plural transmission lines 1-3 almost simultaneously in this way, even if a fault takes place in any of the transmission lines, the message is transferred to an object subsystem without a delay.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a communication control system for a highly reliable local area network (hereinafter referred to as LAN), and in particular, the present invention relates to a communication control system for a highly reliable local area network (hereinafter referred to as a LAN), and in particular for a communication control system for a plurality of sub-subs connected by multiple LAN transmission paths. A highly reliable LAN communication control method that includes systems.

[Conventional technology]

In the conventional highly reliable LAN communication control method, multiple subsystems composed of CPU systems are connected to multiple LAN transmission paths, and the sender sends the same message to the receiver through the multiple LAN transmission paths. Among the same messages received by the message identifier assigned to the message by the receiving side, the receiver receives only the first message received and discards the others.

[Problem to be solved by the invention]

In the conventional high-reliability LAN communication control method described above, when sending data from a sending side to a receiving subsystem (hot standby system or parallel operating system) consisting of two or more CPUs, the receiving subsystem At that time, it is necessary to specify the CPU that is operating at that time and send it, and the receiving side must understand the dynamic and temporally changing system configuration control status and consciously change the communication control on the sending side. However, since communication control cannot be unified, communication control in each subsystem becomes extremely complicated.

[Means for Solving the Problems] The highly reliable local area network communication control method of the present invention includes a plurality of subsystems each including at least one CPU system, and a multiplex system that connects the plurality of subsystems in parallel. In a communication control method for a highly reliable local area network having a local area network (hereinafter referred to as LAN) transmission path and each node that can access each of the multiple LAN transmission paths from each of the subsystems, each Means for the CPU system of the subsystem to perform group broadcast communication using each node of the destination subsystem as a group through its own node and each node of the destination subsystem; The CPU system of the sending side subsystem sends the same message to each of the LAN transmission paths at the same time when sending to the receiving side subsystem, and the receiving side subsystem A function that discards messages other than the first received message among the same messages received by the message identifier assigned to the message by the system, and a response message after the received message is processed by the CPU system of the receiving subsystem. to the transmitting side subsystem through each of the LAN transmission paths, and the transmitting side subsystem completes the communication using the message received first among the response messages.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a highly reliable LAN communication control method according to the present invention.

FIG. 1 shows an example in which communication is performed between dual-configured subsystems connected to a LAN with triple transmission paths.

As shown in FIG. 1, subsystem A includes a CPU system A1
It is a parallel processing system with dual thread input and single output, consisting of a CPU system A2 and a CPU system A2, and both systems are connected to the CPU systems Bl and B2 owned by subsystem B through their respective LAN transmission lines 1 and 2.3. A data link can be established.

On the other hand, subsystem B consists of CPU system B1 and CPU system B2.
It is a hot standby system consisting of

First, the LAN environment definition before operation in each subsystem will be explained.

In Figure 1, CPU system A1 and CP of subsystem A
U-system A2 refers to all nodes owned by subsystem A.
and node A21. Node A12 and node A22. Node A13 and node A23 are connected to LAN transmission path 1. L.A.
N transmission line 2. The LAN environment is defined so that the same group can perform group broadcast communication on the LAN transmission path 3.

On the other hand, CPU system B1 and CPU system B2 of subsystem B are node B11 and node B owned by subsystem B.
21. Node B12 and node B22, and node B13 and node B23 are connected to LAN transmission line 1. LAN transmission line 2
．． The LAN environment is defined so that the same group can perform group broadcast communication on the LAN transmission path 3.

Next, a transmission path management method for setting a logical path will be explained.

In Figure 1, CPU system Al and CP of subsystem A
The U-system A2 manages communication paths corresponding to the respective LAN transmission lines 1, 2.3 for the subsystem B, and each node B11, B12, . B13. B21
．． B22. A logical path passing through B23 can be set.

In this embodiment, CPU system Al, CPU system A2. CPU system B
l. The CPU system B2 manages logical paths with the communication destination subsystems for each LAN transmission path 1, 2.3, and the first node and logical path are set on each LAN transmission path 1, 2.3. When the message is established, it is managed as a communicable state, and the message identifier to be given to the message is initialized to the sending/receiving path of the communication path. With the above steps, communication becomes possible. The setting of this logical path can be performed asynchronously with communication both when recovering from a communication path failure and during normal operation.

FIG. 2 is an explanatory diagram of communication information showing the message format of this embodiment.

Next, the operation when communicating from subsystem A to subsystem B using an inter-subsystem message having the message format shown in FIG. 2 will be described.

The master system CPUAl of subsystem A sends the transfer data to subsystem B.

A message identifier and a message format identifier are assigned. Then, the state of the transmission path to subsystem B is checked, and the transmission path is sent to subsystem B via the LAN transmission paths 1, 2.
Nodes Bll, B12. B13, B21. B22. B
Messages are sent to 23 groups almost simultaneously. Here, in subsystem B on the receiving side, the CPU
It is assumed that system B1 is a mask system and CPU system B2 is a standby system, and in subsystem A on the sending side, CPU system Al is a mask system and CPU system A2 is a slave system.

The CPU system B1 is a node B11. Receives the received message through node B12 and node B13, compares the message identifier given to the message with the message identifier of CPU system B1, selects the message that arrives first, and determines the message identifier. Change according to certain rules. Note that the sending side stores the sent message identifier and destination subsystem name, and starts a monitoring timer.

Then, after processing in its own system, the receiving side sends a response message to which the message identifier of the received message has been added, checking the transmission bus status to subsystem A, and transmitting it through the LAN transmission paths 1, 2.3 that can be sent. Nodes A11, A12 . of subsystem A on the sending side. Al1. A21. A22
．． A response message is sent to group A23 almost simultaneously.

On the other hand, on the receiving side, the same message is received from another LAN transmission path, but this message is checked using the updated message identifier and is discarded because it does not match.

Therefore, a response message was sent to subsystem A.

The message is notified to both the CPU system A1, which is a mask system, and the CPU system A2, which is a slave system.

Based on the message identifier in the response message and the response subsystem name, the mask CPU system Al stops the monitoring timer, and further performs matching processing between the mask system and the slave system according to the communication method within the subsystem A.

Although the same response message is sent from each LAN transmission path, the subsystem A side adopts the first message notified and discards the others.

Perform retransmission.

Note that if the specified number of retransmissions fails, if a hardware failure or the like is detected, or if the communicable state is canceled, the communication path is set to be in a communicable state.

〔Effect of the invention〕

As explained above, the highly reliable local area network communication control method of the present invention transfers the same data to multiple transmission paths almost simultaneously, so there is no possibility that a failure will occur in any of the transmission paths. Even if a communication path fails, the message can be transferred to the target subsystem without delay, and even if a failure occurs in the communication path, the communication bus is simply not used, so the control method is simple and the local network system Since the sending side can send data without being aware of the system status of the subsystem on the receiving side or the fluid state of system configuration control, control is simple and the manufacturing man-hours for local network systems can be reduced. It has an effect.

All, A12. A13. A21. A22. A23゜B
l 1. Bl 2. Bl 3. B21. B22. B2
3...Node.

Claims (1)

[Claims] A plurality of subsystems consisting of at least one CPU system, multiple local area network (hereinafter referred to as LAN) transmission lines connecting the plurality of subsystems in parallel, and multiple said LA from each of said subsystems.
In a communication control method for a highly reliable local area network that has nodes that can access each of N transmission paths, the CPU system of each subsystem connects to the other subsystem through its own node and each node of the other subsystem. means for performing group broadcast communication in which each node is grouped, and means for each subsystem to manage a communication path state to a communication destination,
When the CPU system of the sending subsystem transmits to the receiving subsystem, it sends the same message to each of the LAN transmission paths at the same time, and the receiving subsystem sends the same message to each of the LAN transmission paths using the message identifier assigned to the message. A function to discard messages other than the first message received among the messages, and a function to process the received message in the CPU system of the receiving subsystem and then send a response message to the sending subsystem via each of the LANs.
A communication control method for a highly reliable local area network, characterized by having a function of transmitting data almost simultaneously through a transmission path, and completing communication using the first message received by a subsystem on the sending side among response messages.