JPH05158840A - Frame relay communication processing method - Google Patents

Abstract

(57) [Summary] [Object] According to the present invention, a plurality of processors are connected to a duplicated communication control device that executes communication processing, and the communication control devices of each system are connected to each other by a network such as a duplication ring. Frame relay communication processing method in the multi-processor communication control system
It is an object of the present invention to enable communication between processors connected to a communication control device in which a failure has occurred, even if a failure occurs in each different communication control device in the duplex ring. When a communication control device corresponding to a different processor in a communication control system of each system fails and direct communication cannot be performed by the communication system in each system between the processors, the communication control device fails For communication from one processor to the other processor corresponding to, the frame is relayed from one system on the normal communication control unit side to the other system via the processor whose communication control units on both systems are normal. Configure to do.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a redundant communication control device in which a plurality of processors execute communication processing.
The communication control device of each system relates to a frame relay communication processing system in a multiprocessor communication control system which is mutually connected by a duplicated loop or ring.

In recent years, a method of connecting a plurality of processors via a high-speed ring bus or loop bus (hereinafter, simply referred to as a ring bus) to perform load distributed processing has been used. The ring bus is provided with a communication control device connected to each node at each node, and communication between the processors is performed by the ring bus via a communication terminal device. Such a ring bus is duplicated to maintain reliability, and each processor is connected to a corresponding communication control device on the ring bus of each system. Also, even if each ring bus is duplicated and a failure occurs, communication between other normal communication control devices is possible.

In such a communication control system between multiprocessors, when a failure occurs in a communication control device corresponding to different processors of both systems in a duplicated ring bus, communication between those processors becomes impossible,
The improvement is desired.

[0004]

2. Description of the Related Art FIG. 7 is an explanatory view of a conventional example. In FIG. 7, PE is a processor (Processor Element), CCU
Represents a communication control unit, and each processor PE processes a distributed load, and constitutes a system that executes a set of functions as a whole. Specifically, each processor corresponds to, for example, a processor that dispersively processes and controls the respective communication paths of the exchange.

The plurality of processors PE0 to PEn are provided with duplicated CCUs of 0 system and 1 system, and the CCUs of each system are 0 system ring bus (shown as 0 system ring) and 1 system, respectively.
Communication is performed by connecting with a dual high-speed LAN called a system ring (also referred to as a 1-system ring). This high-speed LA
N is FDDI (Fiber Distributor), which is being standardized by ISO (International Standards Organization) and ANSI (American National Standards Institute).
Buted Data Interface) 100 Mbps optical LAN.

The 0-system ring and the 1-system ring have the same structure, and the 0-system ring will be described as a representative. A ring is a P ring (Primary Ring) that transfers data in one direction.
And an S-ring that transfers data in the opposite direction (Seconda
ry ring), the P ring is used as the active system, and the S ring is used when the P ring fails. As another usage pattern, the P ring and the S ring may be selectively used according to the type of information to be transmitted.

In the ring bus having such a structure, when a failure occurs in the CCU or the transmission path in the ring bus, the P ring and the S ring are connected in the CCU in which the failure occurs or the CCU at both ends of the transmission path. By forming a loopback path, each processor can communicate with each other via another CCU, except for the position where the failure has occurred. In the example of FIG. 7, if a failure occurs in CCU0 of the 0-system ring, loopbacks as shown in the figure are formed between CCU0 and CCU1 of the 0-system ring and between CCU0 and CCU0, respectively, and CCU1 to CCU1.
A circulation path by the S ring and the P ring is formed between the CUns, and the communication between the CCUs operating normally can be maintained. On the other hand, the processor PE0 connected to the CCU0 in which the failure has occurred communicates with other processors, which have been performed via the 0-system ring until then, to the CCU of the 1-system ring.
It can be realized by using 0.

High reliability of the inter-multiprocessor communication control system can be achieved by the dual structure of the ring bus formed by the dual communication control device and the ring bus itself.

[0009]

In the conventional system shown in FIG. 7, when a failure occurs in one communication control unit of the 0-system ring, the failure point of the 0-system ring is different from that of the 0-system ring.
When a failure occurs in one of the communication control units of the system ring (duplication failure occurs), the processor connected to the communication control unit in which the system 0 ring failure occurs and the communication control unit in which the system 1 ring failure occurs It was not possible to communicate directly with the processor connected to.

Explaining in a concrete example, a failure occurs in CCU0 in the 0-system ring of FIG.
The case where a failure occurs in CUn is taken as an example. In this case, 0
In the system ring, CCU0 fails and CCU1 and C
A loopback loop is formed between CUn,
In the system 1 ring, a loopback loop is formed between CCU0 and CCUn-1. In this case, when the processor PE0 communicates with the processor PEn (and vice versa), even if the processors PE0 and the processor PEn try to communicate with each other using the rings of other systems, they cannot communicate because the loopbacks are formed respectively. was there.

According to the present invention, even if a failure occurs in each different communication control device of a redundant ring configured by connecting a redundant communication control device, an interprocessor connected to the communication control device in which the failure occurs It is an object of the present invention to provide a frame relay communication processing method in a multiprocessor communication control system that enables the communication of the above.

[0012]

FIG. 1 illustrates the principle of the present invention. In FIG. 1, 1 is a 0-system ring bus that is duplicated with a P-ring and an S-ring, 2 is a communication control unit (CCU00-CCU0n) provided on the 0-system ring bus and connected to processors PE0-n, 3 respectively. Are processors (PE0 to PEn) that perform individual processing, 4 is a 1-system ring bus that is duplicated like the 0-system ring bus, and 5 is provided on the 1-system ring bus and connected to processors PE0 to n, respectively. Communication control unit (CCU10
~ CCU1n).

According to the present invention, when the 0-system and 1-system ring buses are duplicated and a failure occurs in the 0-system communication controller and the 1-system communication controller corresponding to different processors, the communication in which the failure occurs The communication between the processors connected to the control device is such that both the 0-system communication control device and the 1-system communication control device relay the processors that are normal and transmit / receive frames between the rings.

[0014]

Operation: One of the communication control devices 2 on the 0-system ring bus 1,
For example, one of the communication control units 5 on the 1-system ring bus 4 corresponding to a processor different from the communication control unit in which the fault occurs in CCU00 and the 0-system, for example, CCU1n
It is assumed that there is a failure. In this case, in the 0-system and 1-system ring buses 1 and 4, communication can be performed by forming a loopback in the communication control devices on both sides of the communication control device in which the failure has occurred, as in the conventional case. At this time, the processor PE0 connected to the communication control unit CCU00 having a failure in the 0-system ring bus 1 is connected to the processor P0 connected to the communication control unit CCU1n having a failure in the 1-system ring bus 4.
It shall communicate with En.

In this case, the processor PE0 transmits from the 1-system ring bus 4 having the normal communication control unit CCU10 a frame addressed to the communication control unit CCU0n connected to the partner processor PEn on the 0-system ring. This frame is sent to the other communication control unit CCU11 through the 1-system ring bus 4 (loop formed by loopback). When the communication control unit CCU01 of the other system connected to the processor PE1 corresponding to the own communication control unit is normal, the other communication control unit CCU11 takes in the frame and transfers it to the higher-order processor PE1 of the own communication control unit. .. The processor PE1 sends out the frame to the ring bus of the other system from the communication control unit CCU01 of the other system on the opposite side to the time when the frame is received. In the other system's ring bus (system 0 ring bus 1), each communication control device performs capture control of the frame destined for itself, and in this example, the communication control device CCU1n receives the frame destined for itself and the upper processor Transfers its contents to PEn. In this way, communication from PE0 to PEn is performed. Communication in the opposite direction (communication from PEn to PE0) is executed in the same manner.

[0016]

FIG. 2 is a system configuration diagram of an embodiment, FIG. 3 is a processing flow chart of a processor of a transmission source, FIG. 4 is a diagram showing a frame format, and FIG. 5 is a frame transmission source processor, a relay processor and a final reception destination processor. 6 is an example of frame data of a relay frame.

In FIG. 2, 1 to 5 are circuits or devices similar to the same reference numerals in FIG. 1, 1 is a 0-system ring bus composed of a primary ring (1P) and a secondary ring (1S), and 2 is a 0-system. Communication control device, 3 is a processor connected to each duplicated communication control device, 4 is a 1-system ring bus composed of a primary ring (4P) and a secondary ring (4S), and 5 is a 1-system communication control device. ..

Specifically, this system can constitute a control device of a digital exchange, for example, PE0.
Can correspond to MPRs (management processors) and PE1 to PEn to CPR1 to CPRn (Call Processors for controlling each distributed network).

In FIG. 2, when the communication control unit CCU0 on the 0-system ring bus 1 fails and the communication control unit CCU on the 1-system ring bus 4 fails, the processor PE0 communicates with the processor PEn. The operation in the case of performing will be described with reference to FIGS.

When the processor PE0 communicates with PEn, if a failure (system double failure) occurs in the communication control unit on both the 0-system ring bus 1 and the 1-system ring bus 4 (30 in FIG. 3), the processor PE0 Detects that direct communication to PEn via the 1-system ring bus is not possible.
1). Next, it is detected that direct communication with all other PEs via the 0-system ring bus is impossible (at 32). This allows PE0 to be used regardless of whether the 0-system / 1-system ring bus is used.
Recognizes that the direct communication between the and PEn is impossible (at 33). Next, one PE (in this case, PE1) is designated as the relay communication from the processors PE which can communicate at present (34), a relay frame for PEn is sent to PE1, and the relay processing is performed by PE1. Request (the same 3
5).

Here, FIG. 4 shows a frame format (general type) used when performing the relay processing. As shown in FIG. 4, a frame used for communication between processors has a start delimiter (SD) at the beginning, followed by a frame control (FC), a destination address (DA), a source address (SA), a network header ( NW Header), information (Information), frame check sequence (FC
S), end delimiter (ED) and frame status (FS)
Composed of and.

In this frame, a network header (NW Header) is a field that defines control information for relay processing provided by the present invention. The relay processing is
Either 0-system ring bus to 1-system ring bus or 1-system ring bus to 0-system ring bus, 0-system ring bus to 1-system ring bus, and 1-system ring bus to 0-system ring bus It does not include one-time relay processing.

As shown in detail in FIG. 4, the network header (NW Header) is provided at the beginning with a network frame type (NWT) indicating the type of the network frame, which has the following types.

NWT = 0: direct communication frame NWT = 1: relay communication request frame NWT = 2: relay communication frame Next, the source ring number indicating the ring number of the frame issuer (0 system / 1 system ring bus distinction) (SR), source CCU address (S
A), a destination reception ring number (DR) that represents the ring number of the final reception destination (discrimination of 0-system / 1-system ring bus), and a destination CCU address (DA) that represents the final reception destination CCU.

In FIG. 2, the processor PE0 is P
When communicating to En, the communication control unit fails in both system 0 ring bus 1 and system 1 ring bus 4 (system double failure)
When PE occurs, a relay frame is sent from PE0 (see 35 in FIG. 3), and each processor performs relay processing.

FIG. 5 shows a processing flow in each processor (PE0, PE1, PEn) related to the relay frame in that case. The processor PE0 sends a frame to the CCU via the CCU0 of the 1-system ring bus (called ring 1).
1 (50 in FIG. 5). This frame is a relay frame in the format shown in FIG. 4, and its frame data is shown in FIG. 6 with DA = CCU1.
(Specify CCU corresponding to the relaying processor), SA
= CCU0, NWT = 1 (relay communication request frame)
Then, the ring number 1 of the issuing source and the CCU address 0, and the ring number 0 and the CCU address n of the final receiving destination are set.

This frame passes through ring 1 to the CCU
When it is sent to No. 1 and received by PE1, PE1 checks its NW header, and if it is found that it is a relay communication request frame from PE0, it executes a relay process because it is not a direct communication frame (at step 51).

Then identify the DR / DA of the frame,
Since the frame is addressed to CCUn of ring 0, the frame is transmitted to CCUn of ring 0 (at 52). The frame data for this frame is shown in Figure 6, where DA =
It includes data such as CCUn, SA = CCU1, NWT = 2 (relay communication frame).

This frame is received at PEn via CCUn on ring 0. PEn checks the NW header and recognizes it as a relay communication frame from PE0 via PE1 (at step 53). The PE1 performs a process corresponding to the information loaded in the received frame. When PEn sends a response frame for this frame to PE0, it sends back a relay frame via PE1 by the route opposite to that of reception.

That is, PEn sends a frame to PE1 via Ring 0 (from CCUn to CCU1),
It requests relay processing to PE0 (at 54). The frame data for this frame is shown in FIG. PE
1 receives this from CCU 1 on ring 0, it receives N
The W header is checked, and since it is a relay communication request frame from PEn and not a direct communication frame, relay processing is executed (55). That is, PE1 looks at the DR / DA of the NW header and, since it is a frame addressed to CCU0 of ring 1, creates a frame for relay communication and transmits it to CCU0 via ring 1 (step 56). Frame data of this frame is shown in FIG.

This frame is sent to PE0 via ring 1.
When the packet is received at, the NW header is checked to recognize that it is a relay communication frame from PEn to PE0 (at step 57). That is, the SA field (= CCU1) of the frame and the SR of the NW header = ring 0 /
SA = CCUn, PEn to PE1 are ring 0 Cs
It can be seen that the frames from PE1 to PE0 are relayed via CCU1 and CCU0 of the ring 1 via CUn and CCU1.

Through the above series of processing, PE0 to PEn
To PE0 and PEn to PE0 can be realized. Thereafter, the relay communication between PE0 and PEn is executed by the same processing procedure, and the online processing is continued until the failure is recovered.

The above-described relay communication processing according to the present invention can be appropriately performed not only for the duplicated ring bus but also for other types of networks such as a loop bus.

[0034]

According to the present invention, in the case of a double failure in a network such as a ring formed by each of the duplicated communication control units, a frame with a processor having a communication control unit in which both systems are normal is used. By relaying and transmitting / receiving, it is possible to prevent communication down (a state in which communication is impossible) between processors having a communication control device with a failure, and it is possible to realize fail software. Further, the relay processing is realized by online software, and the protocol of N layer (network layer) can be used.

[Brief description of drawings]

FIG. 1 is a principle configuration diagram of the present invention.

FIG. 2 is a system configuration diagram of an embodiment.

FIG. 3 is a processing flow chart of a processor of a transmission source.

FIG. 4 is a diagram showing a frame format.

FIG. 5 is a process flow chart of a frame source processor, a relay processor, and a final destination processor.

FIG. 6 is an example of frame data of a relay frame.

FIG. 7 is an explanatory diagram of a conventional example.

[Explanation of symbols]

1 Redundant (1P, 1S) 0-system ring bus 2 Communication controller on 0-system ring bus (CCU0
0 to CCU0n) 3 processor (PE0 to PEn) 4 dual (4P, 4S) 1-system ring bus 5 communication controller on 1-system ring bus (CCU1)
0 to CCU1n)

Claims (2)

[Claims] 1. A plurality of processors are each connected to a duplicated communication control device that executes communication processing, and the communication control devices of each system are connected to each other by a network such as a duplicated ring. When a failure occurs in the communication control device, the communication control systems of the respective systems differ in the multiprocessor communication control system that forms an intercommunication path by folding back the duplicated network before and after the communication control device in which the failure occurs. When a failure occurs in the communication control device corresponding to the processor and direct communication by the communication system in each system between the processors becomes impossible, one processor corresponding to the communication control device in which the failure has occurred to the other processor Communication from one system on the normal communication control unit side to the normal communication control unit on both systems. A frame relay communication processing method in a multiprocessor communication control system, characterized in that a frame is relayed to another system via a processor. 2. The relay frame according to claim 1, wherein communication from one processor corresponding to the communication control device in which the failure has occurred to the other processor is performed by using a relay frame. A network frame type for displaying whether it is for relay communication, a system number of a issuing source from one system of a normal communication control device, a communication control device number, a system number of a final receiving destination, and a communication control device number. Frame relay communication processing method for multiprocessor communication control system.