JPH04239831A - Inter processor backup system - Google Patents

Abstract

PURPOSE:To improve the reliability allowing a network control processor to possess exclusive network configuration information and allowing a network management processor to control the information corresponding to any network system. CONSTITUTION:A network management processor 1 and five network control processors 2a-2e connecting to a common control bus form 3 network systems implementing jobs A, B, C. Each of the processors 2a-2e is connected to each of the network systems by matrix type control line changeover switches 3a, 3b. Each of the processors 2a-2e stores network configuration information of each system and the processor 1 controls each system accordingly. In this embodiment, number of processors of the active system and the standby system is selected to be 3:2, and this ratio is optionally set by taking the importance and the economical condition into account.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mutual processor backup system, and more particularly to a mutual processor backup system for efficiently preventing system down when a network control processor that processes communications in a network system becomes impaired. .

0002

2. Description of the Related Art In current large-scale network systems, as shown in FIG. The network control system 8 is in charge of this part. In this way, the processing capacity of the entire network is increased by freeing the processing of the host computer 4 from data communication and making it specialized for data processing.

[0003] Against this background, the network control system 8 in a large-scale network system is positioned as the core of the network system, and is required to have high operability. In order to achieve this high availability,
Conventionally, backup methods for when the processor of the network control system 8 fails have been discussed.

A major conventional example is a duplex system as shown in FIG. This method uses two network control processors of the same type, one in active system and one in standby system.
The network control system 8 is configured to prevent the network system from going down by switching the line to the standby processor when the active processor fails.

[0005]

[Problems to be Solved by the Invention] The conventional duplex system described above requires two network control processors, one for active system and one for standby system, for one host computer, which makes it economically expensive. There are drawbacks to this. Furthermore, when a failure occurs in both the active system and the standby system, there is no way to back up the system, and system down of the network system is inevitable. In particular, when the host computers of a plurality of network systems are located in one computer center, it is not necessarily considered an efficient method to make each system a duplex system. However, since each network system has a different configuration and the network control processors are not necessarily the same, there is a problem that it is not possible to simply use a common standby processor.

The object of the present invention is to eliminate the above-mentioned problems and
An object of the present invention is to provide a mutual processor backup method that can commonly back up multiple network systems.

[0007]

[Means for Solving the Problems] The mutual processor backup system of the present invention provides an N+
m (m is a positive integer) network control processors, a network management processor that monitors and manages the operating status of each of the network control processors, and a line switch that can switch and connect each of the network control processors to each of the network systems. and a switch, wherein each of the network control processors is instructed by the network management processor from a non-volatile storage area storing network configuration information of the N network systems in a format suitable for each of the network control processors. The network system is configured to include a network entry means for entering the network system after reading the network configuration information of the network system.

[0008]

Embodiments Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention.

The mutual processor backup system of this embodiment, as shown in FIG.
The communication control system consists of one network management processor 1 connected to a common control bus, five network control processors 2a, 2b, 2c, 2d, and 2e, and a system console 7, and is a matrix type system. The line changeover switches 3a and 3b allow each network control processor 2i (i=a to e) to be connected to any network system.

[0011] Five network control processors 2a~
2e, network control processors 2a, 2b,
The three network control processors 2c are installed as the current processors (current A, current B, and current C) that configure each network system, and the other two network control processors 2d and 2e
is a standby processor for common backup (standby 1
, standby 2).

Next, referring to FIGS. 2 and 3, the network management processor 1 and the network control processor 2i
The internal functional configuration of is explained below.

As shown in FIG. 2, the network management processor 1 includes an operating status control section 11 and a message control section 1.
2, a functional section consisting of a command control section 13 and a processor control section 14, and a data section consisting of a network control processor status management section 15.

As shown in FIG. 3, the network control processor 2i has a functional section consisting of an operating status control section 21, a command control section 22, and a processor control section 23;
It consists of a data section consisting of a network configuration information save area 24 and a network configuration information storage area 25.

The operating status control units 11 and 21 implement an operating status checking function and a faulty processor identification function. The operation status control unit 11 of the network management processor 1 transmits an operation status check message to the operation status control units 21 of all the network control processors 2i. After transmission, all message statuses 16 in the network control processor status management unit 15 are updated to "waiting for response message." Each network control processor 2i
Upon receiving the operation status check message, the operation status control unit 21 immediately returns a response message with the processor number of its own processor attached to the operation status control unit 11. When the operating status control unit 11 receives the response message, it checks the attached processor number and updates the message status 16 corresponding to the processor number to "response message received." If the message status 16 remains "Waiting for response message" even after a certain period of time has passed, it is recognized that a processor failure has occurred, so the status 17 of that processor number is updated from "normal" to "fault". . Then, the message control unit 12 is notified of the processor number that caused the failure and the current type 18 information.

The message control unit 12 outputs a message displaying the contents on the display screen of the system console 7 every time an event change in the network such as a processor failure occurs, and notifies the system operator of the event. I do.

The command control units 13 and 22 transmit and analyze instructions from the operator. This realizes the function of instructing the standby processor to back up when the active processor fails.

The processor control units 14 and 23 manage the state of their own processors. In particular, when the network control processor 2i receives a backup instruction notification from the command control unit 22, it analyzes the type of network system (current A to C) of the processor to be backed up, and then performs the backup from the network configuration information save area 24. Network entry is realized by finding the network configuration information of the target system, storing it in the network configuration information storage area 25, and then initially reloading the necessary programs. The network configuration information save area 24 is a nonvolatile storage area that stores network configuration information of each network system in the format of its own processor.

Referring now to FIG. 4, the flow of backup of the network control processor in this embodiment will be explained.

In the initial state T1, the network control processors 2a, 2b, and 2c are set as active A, B, and C, and the network control processors 2d and 2e are set as standby 1 and standby 2.

Assume that at a certain point in time T2, the network control processor 2b, which was operating as the active B, becomes faulty. The operating status control unit 11 of the network management processor 1 that has detected the failure of the network control processor 2b sets the status 17 of the processor number “2b” in the network control processor status management unit 15 as “failure” and then sends a message to the network control processor 1. Network control processor 2b was operating as active B for 12.
Notify you that there is a problem. In response to this, the message control unit 12 outputs a failure message for the working B (12b) on the display screen of the system console 7.

[0022] Time T when the operator recognized the failure of the network control processor 2b operating as the active B
3, the network control processor 2d, which was operating as a standby processor (standby 1), is instructed to join the active B network system as a backup. Note that since the processor to be backed up can be arbitrarily selected, the operator only has to issue a backup instruction to the standby processor with the lowest usage rate at that time.

This backup instruction is transmitted from the command control section 13 of the network management processor 1 to the command control section 22 of the network control processor 2d, and then the processor control section 23 is notified of the backup instruction. After analyzing that the type of the network system to be backed up is current B, the processor control unit 23 saves the current B from the network configuration information save area 24.
Find network configuration information related to the network configuration information and set it in the network configuration information storage area 25. Then, by loading the initial program again, the network control processor 2d enters the network as the working B at time T3.

Next, suppose that at a certain point in time T4, the network control processor 2d operating as the working B becomes malfunctioning, and it becomes necessary to back up the working B again. At this time T5, the operator instructs the network control processor 2e to back up the current B.

At the time T6 when the fault in the network control processor 2b has been recovered, the current A, current B, and current C are all operating normally, so the operator can
, the network control processor 2b is instructed to stand by as standby 1.

[0026] According to the above flow, even though the working B has failed twice, it has been sequentially backed up using the mutual processor backup method, so that the system failure of the B business network system can be prevented.

In the above embodiment, the ratio of the number of active and standby processors is 3:2, but this ratio takes into consideration the failure rate of hardware devices, the importance of the network system, and economic conditions. The network system can be configured flexibly depending on the conditions because it can be set as desired. Furthermore, although each network control processor has been described as having its own network configuration information save area, it may be common to processors of the same type and standard.

[0028]

As explained above, in the mutual processor backup method of the present invention, all the network control processors have the network configuration information of each network system exclusively, and the network control processor can handle any network system. Under the control of the management processor, it is possible to configure an N to m (N and m are both positive integers, N>m) backup system. In particular, when m is 2 or more, it is possible to flexibly deal with double failures of the network control processor, and there is an effect that high operability and reliability can be economically provided to the network system.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention.

FIG. 2 is a functional configuration diagram of a network management processor in this embodiment.

FIG. 3 is a functional configuration diagram of a network control processor in this embodiment.

FIG. 4 is an explanatory diagram showing an example of a backup flow in this embodiment.

FIG. 5 is a block diagram of an example of a conventional duplex network configuration.

[Explanation of symbols]

1 Network management processor 2, 2i (i=
a to e) Network control processors 3a, 3b Line changeover switch 4 Computer 5 Distributed processing processor 6 Business terminal 7 System console 8 Network control system 11, 21 Operating status control section 12 Message control section 13, 22 Command control section 14, 23 Processor control part

Claims (1)

[Claims] Claim 1: For N (N is an integer of 2 or more) network systems, N+m (N+m) that share communication processing.
(m is a positive integer) network control processors, a network management processor that monitors and manages the operating status of each of the network control processors, and a line changeover switch that can switch and connect each of the network control processors to each of the network systems; each network control processor has a network of network systems instructed by the network management processor from a non-volatile storage area storing network configuration information of the N network systems in a format suitable for each network control processor. A mutual processor backup method characterized by comprising a network entry means for entering the network system after reading configuration information.