JPH06101740B2 - Network configuration control method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data exchange network having a star ring structure, and more particularly to a method for separating a failure point (configuration control), which is suitable as a countermeasure against a failure when a transmission line is abnormal.

[Background of the Invention]

The network structure to which the present invention is directed is the star ring structure shown in FIG. In FIG. 1, 1 (A,
…, G) is the station (hereinafter referred to as ST1), 2 (A, B)
Is the station that controls the configuration of the network (hereinafter,
In order to distinguish from ST1, the management ST is used.) 3 is a concentrator,
4 is a transmission line, 5 (A, ..., G) is a port that is a unit of disconnection, 6
Is a signal line for port disconnection, and 7 is a data flow. Especially in Fig. 1, one ST1 per port
However, multiple ST1s may be connected to one port. Note that FIG. 1 shows a configuration in which the port 5-F is separated (BYPASS ... Bypass).

Conventionally, there is known a method described in the following document which is used as a countermeasure against a failure of a network having the above structure.

1. Local Area Network IE on May 8, 1982
Day D'Abriille, contributed to the EE802 Standardization Committee
Draft by Andreuse and G. S. Schultz “Towan for Local Area Networks”
Ring Architecture "(Contribution of workin
g papers to IEEE Project 802 on Local Area Netwo
rks, March8,1982 “A Token Ring Architecture for Lo
cal Area Networks "By DWAndrews, GDSchultz) 2. JP-A-58-117750 (i) According to the method of JP-A-58-117750, each ST has an identifier of the port (port) when connecting to a network. Number).

(Ii) Each ST sends an abnormality notification frame when the network is abnormal. The frame has a predetermined structure and is transmitted to all STs and STs in the case. The ST and the management ST that have received the frame stop the transmission if the own ST is transmitting the abnormality notification frame. As a result, only one ST or management ST immediately downstream of the failure point continues to send the abnormality notification frame.

(Iii) The port number is stored in the abnormality notification frame. If the port number is uniquely assigned to the entire network, the management that received the frame
The ST can determine whether the fault exists under the concentrator of the own ST. If it is determined that the port is under its own jurisdiction, the faulty port can be specified.

The already known method has been described above. However, the above method has the following problems.

1. A special mechanism (device) for giving a port identifier to ST is required.

2. If the port identifier is set incorrectly, the failure (port) cannot be bypassed.

[Object of the Invention]

An object of the present invention is to provide a method for disconnecting a failure point (failure port) in a star ring network, and in particular, a method for achieving the above object without using a mechanism for giving the port number identifier. To do.

[Outline of Invention]

As a countermeasure against the failure of the Star Ring configuration network, the problem in performing configuration control is: 1. Who (where the concentrator) performs configuration control, 2. Where (port) is disconnected .

According to the present invention, it is not necessary to give the port identifier to the ST as in the prior art, and the first problem can be solved as follows.

In other words, manage all concentrators that can control the configuration.
In ST, a new frame (configuration control frame, hereinafter, different from the abnormality notification frame when the network is abnormal)
RCF) is sent. Administrative ST that cannot receive RCF
The concentrator under the jurisdiction is a concentrator that should take countermeasures against failures, and a concentrator that should perform configuration control. This is because the management ST is the management ST located downstream of the fault point. Conventionally, there is the following solution to the second problem. By disconnecting the port by trial and error, the transmission source ST address of the abnormality notification frame is changed.
By observing this, the actual failure location can be specified. This method does not assume the existence of RCF. In other words, RCF is not used to identify the faulty port.

This trial-and-error method of disconnecting a port, that is, a method that does not use RCF, basically has the following problems. Consider a case where there are a plurality of concentrators. First, since there is no concern about which concentrator should start the configuration control (trial of disconnection), there is a problem that the configuration control must be performed by all the concentrators in the network. In contrast,
In the method using the RCF as in the present invention, only one concentrator including the faulty port can control the configuration.

The present invention particularly uses RCF to identify a faulty port. That is, instead of monitoring the fluctuation of the ST address of the transmission source of the abnormality notification frame, it is monitored whether the RCF returns. The action (in this case, disconnection) is suddenly performed before searching for the location of the faulty port. If the fault location is cut off, the ring-shaped transmission line is "closed" and the RCF loops back through the ring. Receiving the RCF means that the troubleshooting is complete. The point of failure can be identified at this point.

Example of Invention

 Examples of the present invention will be described below.

FIG. 2 shows the structure of a frame used in this embodiment. 10 is
Start Delimiter (SD) indicates the start of frame, 11 is F
rame Control (FC) indicates the frame type, 12 is Dest
Destination address is indicated by ination Addres (DA), 13 is Source
Address (SA) indicates the source address, 14 is Informat
ion (I), 15 is Frame Check Sequence (FCS), 16 is En
d Delimiter (ED) indicates the end of the frame. Incidentally, I may be omitted. Also, DA is the broadcast (or management ST group) address to all management STs.

The following frame is used in this embodiment. If the configuration control frame (hereinafter abbreviated as RCF) FC11 is 1-byte information, it can be defined as follows.

FC11 ... 10100001 (RCF) FIG. 3 shows a configuration control flow in the management ST2. In this embodiment, the operation of ST1 will not be described. In the management ST2 that performs configuration control, the faulty port disconnection (bypass) operation is described especially when a port error occurs.

The network is a serial system, and even if one port fails, it becomes an abnormality that affects the communication of the entire system. For example, data transmission timing signal abnormality,
It is reflected in the disappearance abnormality of the token.

The management ST2, which has detected the abnormality, sets the mode in which the RCF is continuously sent to the downstream continuously in step 101, and the R in step 102 is set.
Start the CF reception monitoring timer (RCFWaiting Timer: TWR). The value of TWR may be, for example, 1 second.

Self-management by monitoring RCF reception at step 103
Determine whether the ST should perform configuration control. If the RCF is received, it turns out that there is another management ST upstream from the self management ST for which another configuration control should be performed. Step
The TWR is stopped at 112, the RCF continuous transmission mode is canceled at step 113, and the received RCF is relayed downstream.

If the TWR times out (104) without receiving the RCF,
It judges that there is a faulty port in the line concentrator under the control of the self-managed ST and starts configuration control.

At steps 105 and 106, the port P located at the uppermost stream of the concentrator is bypassed, and at step 107 TWR is started. Step 1 by monitoring the reception of the RCF here.
At 08, it can be determined whether the bypass is valid (ie, the faulty port has been disconnected). If an RCF is received, it is considered valid, TWR is stopped at step 112, and step 11
Use 3 to cancel the continuous transmission mode of RCF.

On the other hand, if RCF is not received and TWR times out in step 109, the bypass is considered invalid, step 110 cancels it, and step 111 sets the identifier N to bypass the next downstream port. To do. The next bypass is the port (P-) which is the next downstream of the most upstream port (P).
1). Hereafter, the operation of steps 106 to 111 is RC
It will be repeated until F is received.

So far, the method of bypassing the uppermost port to the downstream ports one by one has been described. However, the above-described embodiment is not sufficient for multiple failures distributed over a plurality of ports. For these, it is easy to consider a method of sequentially bypassing a plurality of ports, a method of bypassing all ports collectively in advance, and then canceling the bypasses one by one. However, both methods are the same in that RCF is used.

〔The invention's effect〕

According to the present invention, it is not necessary for each ST to be aware of "what port the ST is connected to" even under the conditions that allow the ST to freely connect / disconnect / disconnect the network. There is no problem that failure countermeasures cannot be taken due to erroneous setting of, and there is an effect that physical quantity can be reduced by not requiring a special mechanism (device) for giving a port identifier to ST.

Due to the above effect, it becomes possible to disconnect the port in the abnormal state from the network. In particular, it can be applied to multiple failures dispersed in different ports and concentrators, and has the effect of improving the reliability of the network.

[Brief description of drawings]

FIG. 1 is a diagram showing a network of a Stirling configuration, FIG. 2 is a configuration diagram of a communication frame used in the embodiment, and FIG. 3 is a failure countermeasure (configuration control) processing procedure in a management ST. It is a flow chart.

Claims (2)

[Claims] 1. A plurality of concentrators each including at least one port for connecting a terminal station and a management station having a network configuration control function,
In a network connected by a ring-shaped transmission line having a predetermined signal transmission direction, each management station is caused to send a configuration control frame to the ring-shaped transmission line when a network abnormality is detected and When the configuration control frame is received from the upstream side of the ring-shaped transmission line, the transmission operation of its own configuration control frame is stopped, and the configuration control frame is sent out without receiving the configuration control frame from the upstream side of the ring-shaped transmission line. Change the configuration of the transmission system so that the management station that has continued the operation for a predetermined time performs the configuration control operation for the concentrator under its jurisdiction, and the transmitted configuration control frame can go around the ring transmission path. A network configuration control method characterized by the above. 2. A management station which should perform the configuration control operation is caused to check whether or not a configuration control frame is received while selectively bypassing a port in a concentrator under its control, and the configuration control is performed in the bypass state. The port that could not receive the frame for communication is returned to the connected state, and when the configuration control frame can complete one round in the ring transmission line in the bypass state, the configuration control operation is terminated with the port concerned bypassed. The network configuration control method according to claim 1, characterized in that.