JPS58116840A - Centralized annular type data communication network system - Google Patents

Abstract

PURPOSE:To prevent great deterioration in function even in case of the occurrence of a fault by compensating a functional breakdown in case of fault occurrence which is a fault of a centralized network through an annular type network. CONSTITUTION:A node P1 when to transmit data to a node P2 sends a connection request to a central control node C. The node C checks the in-use state of a bus between the nodes C and P2 and sends a P1-P2 bus connectable signal to the node P1 when the bus is usable, thus allowing the node P1 to start necessary connection control. If the bus between the P1 and P2 is not usable, a bus between nodes P3 and P2, etc., is checked. If a fault ocurs to the node C, the occurrence of the fault is reported to respective nodes P1-P6 to disable a centralized network. Then, the P1-P6 are connected in a looped annular type network for communication.

Description

DETAILED DESCRIPTION OF THE INVENTION (1) Technical field of the invention The present invention relates to a data communication network system,
In particular, the present invention relates to centralized and circular network configurations, and to data communication network systems that enable switching control depending on operational conditions.

(2) Background of the technology Processing is performed between multiple information processing equipment units located at geographically distant locations, such as online data processing systems that include many terminal devices and distributed processing systems that combine multiple computers. Data communication usually takes the form of a centralized network or a ring network.In a centralized network, multiple nodes are connected to one central node (central control node), and five communications This is a network that centralizes exchange and monitoring and control functions. This method has problems with system reliability, as if a failure occurs in the central node or line, system functionality will be disrupted over a wide range of areas.

On the other hand, a circular network is a system in which a plurality of nodes are connected in a loop, and each node is provided with a communication control function. In this method, all the lines linking each node are shared by each node, so if the amount of traffic is large.

It has the disadvantage of frequent waiting states.

(5) Purpose of the Invention The purpose of the present invention is to provide a network system that has a certain degree of durability against failures that occur within the network, and that can maintain relatively high traffic processing efficiency when failures occur. Our goal is to provide the following.

(4) Structure of the invention The present invention solves 11 drawbacks of conventional centralized networks.
The above purpose is achieved by compensating for the function down when X.14I occurs using a circular network.The configuration for this purpose consists of multiple nodes having distributed network control functions and one central control node. , comprising a centralized network that connects the plurality of nodes to the central control node, and a circular network that connects the plurality of nodes in a loop.

Usually only the centralized network is functional, and in the event of a failure or excessive traffic condition on the lines of the centralized network connecting any of the nodes forming the loop to the central control node, the central control The node notifies the nodes on both sides adjacent to the node coupled to the line to incorporate the mate into the centralized network via the ring-shaped network line between the adjacent nodes by their network control functions; , if a failure occurs in the central control node, the central control node notifies all the nodes forming the loop to stop the connection function of the centralized network, and replaces it with the circular network instead. (5) Embodiment of the Invention FIG. 1 is a system configuration diagram of an embodiment of the present invention. In the figure, &P1 to P6 are nodes (hereinafter referred to as general nodes) with which data communication connections are arbitrarily established, and C is a central control node. Moreover, the solid lines connecting each P1 to P6 and C are as follows.

The dotted line that represents a centralized network and connects each P1 to P6 in a loop represents a circular network.

In the normal state of the system, only the centralized network is functioning. In the functioning state of a centralized network, data communication connections between common nodes are controlled by a central control node C using conventional control procedures. Any general node requesting a communication connection issues a connection request to the other node to the central control node C, and the central control node C establishes the connection after confirming that the path to the other node is empty.

FIG. 2 shows connection control when some kind of failure occurs in the path between the general node P2 and the central control node Q. In this case, the ring-shaped network part of P2-Pl or Ps-Pl is functionalized and used, and P2 becomes pl
-a or Ps-G centralized network and connect for one communication.

FIG. 6 shows connection control when a central control node GK failure occurs. In this case, the centralized network is disabled and instead loops between P1 and P6. Communication connections are made in a ring-shaped network.

FIG. 4 is a diagram 70 showing the main TF & continuation control procedure in one implementation row. 1, when node P1 needs to perform data transmission to node P2.

At 2, Pl issues a connection request to the central control node C.

At step 6, node C issues the use status of the path between C and P2 to Ape, and if it is free, it issues a connectable (OK) signal to Pl at step 4. When the node P1 receives the connection enable signal, it performs control in step 5 to establish a connection for data transmission.

Also, if the path between hsc- and c-p2 is not possible or cannot be reversed due to traffic concentration, check the Pl-P2 path in step 6, and if it is usable, connect the P1-P2 bus in step 7. It notifies the enable (OK) signal ttP1 and causes Pl to start necessary connection control. Additionally, at 6, if the Ps-P2 bus is not available, at 8 other suitable methods are checked, such as checking the Ps-P2 path. In the configuration diagram, 10 is an exchange switch section.

11 is the connection control unit, 12 is the obstacle 4! ! 15 is a traffic trend monitoring unit; 14 is a poor connection table;

15 is a priority control table, 16 is a network type switching control unit, and ft is shown.

The exchange switch unit 10 is connected to the interface 7A of each general node P1 to P6. The connection control unit 11 controls the exchange switch unit 100 and other necessary controls in accordance with connection requests or connection release requests from each of the nodes P1 to &.

When the failure detection unit 12 detects a failure inside the central control node or the occurrence of an 884 path, it notifies the network type switching control unit 16. Traffic status supervisor 4I! The section manages the waiting state of connection requests for each node, such as the size of the queue and the connection waiting time for one line, and controls network type switching when Hirapics are excessively concentrated and a path with a long waiting time appears. Notify Department 16.

The connection status table 14 is a table that displays the usage status of paths between each node, and the priority control table 15
is a table that indicates the order of competing connection requests as necessary. The connection control unit 11 performs connection control while checking or updating these tables.

The network type switching control unit 16 selects a detour route or a circular route based on information such as the content of the failure or the traffic concentration path. Select the required node.

Figure 6 shows general notes (P i > cosltllt
This is a vJlt diagram. In the same figure, 17 is the node P1
゜18 is a connection control unit for the centralized network function.

19 is an interface, 20 is a network type switching unit, 21 is an operating status display plug, 22 is a connection control unit for the circular network function, 22a to 22θ are internal processes of the control unit 22, 25 is a connection request signal register, and 24 is a A data buffer 25 indicates a transfer control unit.

In a normal state, only the centralized network connection processing section below the broken line in FIG. 1, consisting of elements 19 to 21, is functionalized. Here, if it is necessary to create the network state shown in FIG. 5 and a format switching signal is sent from the central control node C, the network format switching section 20 will instruct the network state above the broken line in FIG. parts are functionalized. The diagram only depicts the processing of ζ connection requests issued from nodes on the left to nodes on the right.

It is assumed that the annular connection control unit 22 is functionalized and the V raster 25 receives a connection request signal from the node Pi-1. Ii! The connection request signal consists of, for example, a label REG, a destination node name, and a source node name. Connection requests are processed as follows.

First, in order to identify whether the self node is free or not, that is, whether a session is opened or not, the #bad behavior display 7 lag 21 is checked. If the node is vacant, a foreshadowing flag is set in step 22b, and if it is in operation, a connection-disabled signal is sent back to the node Pi-1. If it is vacant,
In step 22C, it is further determined whether the destination node name is the own node.

If it is its own node, it sends a connectable signal to node P1-1 at 22d. Upon receiving the connectable signal, node Pi-1 starts data transmission control.

If the name of one incoming node does not match the own node name, the connection request from node Pi-1 is sent to node P.
Since it is for a node further to the right than i, the *vA* request signal in the register 25 is the node PI-z.
In this case, the node Pi instructs the transfer to
remains operational for possible subsequent data transfer services. However, if the destination node on the right or a node in a different route is operational, A connection failure signal is sent back. if,!
When the I/O signal is received, the previously set status display 7 lug 21 is reset. This unreachable signal is also conveyed to node Pi-IK, where it invalidates the connection request. the! 1. Therefore, the predetermined algorithm pressure.

When the connection request is challenged again and the 0 connection request is successful, data transfer is performed via the data buffer 24 under the control of the transfer control unit 25.

If node P1 is node P1 in the state of Thg2 diagram, node P s+1 is node P
2, and the node P1-1 is connected and configured to be the central control node C. In this case, appropriate communication connections are made between the upper and lower dashed lines in node 17 in FIG. This control is performed by the network format switching section 20.

FIG. 7(a) is an explanatory diagram of the processing procedure when the session between Pt and P2 is released, and FIG. 7(b) is an explanatory diagram of the processing procedure when the central control node C goes down due to occurrence of a failure.

The embodiment described above is merely an example of a possible configuration. Also, the bus between each node.

It is also possible that it is a public line.

(6) Effects of the Invention As described above, according to the present invention, it is possible to realize a data communication network system that is highly reliable and does not suffer from significant functional deterioration even when a failure occurs.

[Brief explanation of the drawing]

FIG. 1 is a network configuration diagram of an embodiment of the present invention. Figure 2 is a diagram of the network configuration when a silk failure occurs. Figure 5 shows the network (a) when the central control node is down.
, [Yes]) are explanatory diagrams of a part of the processing procedure. In the figure, 9 is a central control node, 10 is an exchange switch section, 1
1 is a connection control unit C116 is a network type switching control unit, 17 is a node pi, 18 is a centralized network connection control unit, 20 is a network type switching unit, 22 is a circular connection control unit, 26 is a connection request signal V register, 24 indicate data buffers, respectively. Patent applicant: Fujitsu Limited (Figure 1)

Claims (1)

[Claims] A plurality of nodes having distributed network control functions. a central control node; and a centralized network of nodes coupled to the central control node. The scissors are equipped with a circular network in which multiple nodes are connected in a loop.1 Usually, only the centralized network is functionalized, and a central control node is connected to any one of the multiple nodes to form a core loop. In the event of a failure or excessive traffic condition on a link in a network, the central control node notifies the nodes on both sides that are adjacent to the node connected to the link and uses their network control functions to The nodes are integrated into the centralized network through a circular network line between the nodes, and if a failure occurs in the central control node, the central control node notifies all of the nodes forming a loop. A centralized circular data communication network system, characterized in that the connection function of the centralized network is stopped by notifying the operator, and a circular network is activated instead.