JPH05127937A - Multiprocessor system - Google Patents

Abstract

PURPOSE:To provide a multiprocessor system which designates a substituted processor while a system is operated even after a certain processor having a fault is detached from a system bus and a maintenance tool communicates with the processor so as to execute maintenance if the fault occurs in the proces sor when the maintenance tool communicates with the processor. CONSTITUTION:This system is provided with a maintenance tool 36 having a fault information storage memory 36b storing fault information after fault information on the fault processor is read, a substitution table 36c designating a card substituting the function of a fault card and a command check part 36d detecting that the processor concerned has the fault and designating the substitution card, a general-purpose communication device interface 25 having the same function as the tool and a fault processing processor 14 managing the fault of the system are provided.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multiprocessor system having a plurality of arithmetic processing units.

[0002]

2. Description of the Related Art FIG. 6 is a diagram showing a conventional multiprocessor system disclosed in Japanese Patent Laid-Open No. 64-88677.
In the figure, 51 to 5N are processors that execute application programs, 61 is a system bus that connects the processors 51 to 5N, 14 is a failure processing processor that performs system failure management, and 25 is an external device such as a hard disk. General-purpose communication device I for input and output
/ F and 36 are maintenance tools for sending processing requests to the processors 51 to 5N for performing maintenance on the multiprocessor system, and 62 is a network.

In the above configuration, the processors 51 to 5N
Have the same internal configuration. For example, the processor 51 includes a control CPU 51a and a processor 51.
Self-diagnosis control unit 51b for detecting and checking the failure of
A system bus I / F unit 51d that analyzes and executes a command for performing actual processing, and a command processing unit 51c that executes the command and a control for the processor 51 to communicate with other processors 52 to 5N via the system bus 51e. I have it.

Next, the operation will be described. FIG. 7 is a flowchart showing the operation of the fault processing of the multiprocessor system having the above configuration. Here, the processor 51
This fault processing will be described assuming that a fault has occurred. The maintenance tool 36 is connected to the general-purpose communication device interface 25, outputs a maintenance processing request to an arbitrary processor of the multiprocessor, and displays the response thereof. In the processors 51 to 5N of the multiprocessor configuration, when a failure such as a memory parity error occurs during the execution of the program on the processor 51, the self-diagnosis control unit 51b recognizes the occurrence of the failure by an interrupt or the like (S21). ..

The self-diagnosis control unit 51b determines that the processor 51 is in failure from the cause of this interrupt, and notifies the system bus I / F unit 51d of the cause of failure by a message or the like. By this notification, the system bus I / F unit 51d notifies the failure processing processor 14 of the failure information (S22).
When the failure processing processor 14 receives the notification of the occurrence of a failure from the processor 51, the system bus I of the processor 51 is passed via the system bus 61 so that the processor 51 does not affect the execution of other normal processors 52 to 5N. Instructs the / F unit 51d to disconnect the system bus (S
23).

In response to this instruction, the system bus I / F unit 5
1d disconnects the processor 51 from the system bus 61 (S24). At the same time as this disconnection, the failure processing processor 14 causes the other normal processors 51N to
The processing such as notifying the disconnection is performed (S25).

[0007]

Since the conventional multiprocessor system is constructed as described above, when the maintenance tool communicates with the processor, if the card is in a serious failure state or the like, the communication is impossible. Therefore, there is a problem that the maintenance of the function of the processor cannot be executed.

The present invention has been made in order to solve the above problems, and even when a processor fails during system operation and is disconnected from the system bus, a maintenance tool or general-purpose communication is provided. By providing the device interface with a substitution table showing the card that substitutes the function of the faulty card disconnected from the system bus, the maintenance tool can output the same request to the card that substitutes the function of the faulty card. The purpose is to obtain a maintainable multiprocessor system.

[0009]

In a multiprocessor system according to the present invention, a failure information storage memory for storing information of a processor separated from a system bus due to a failure is used as a general-purpose communication device interface or a maintenance tool. If the general-purpose communication device interface or the maintenance tool has an alternative table that specifies the processor that replaces the function when the processor fails, you want the maintenance tool to request maintenance processing according to the information in the alternative table. When the processor is in a failure state, the request is sent to the alternative card for processing.

[0010]

The multiprocessor system according to the present invention is
By the above means, even when a failure occurs and the processor is disconnected from the system bus, maintenance can be performed by outputting a request to the processor that substitutes its function.
Maintainability is improved.

[0011]

【Example】

Example 1. Hereinafter, Embodiment 1 of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of the present invention, and the same parts as those of FIG. 6 showing the prior art have the same reference numerals.
Reference numeral 61 is a system bus, 51 to 5N are processors connected to the system bus, 41 is an external storage device including a hard disk or the like, 62 is a network, 25 is input / output with the external storage device 41, and with the network 62. A general-purpose communication device interface that performs an interface, 14 is a failure processing processor that performs failure processing of the entire system, and 36 is a processor 51 to 5N, a general-purpose communication device interface 25, and a failure processing processor 14 for performing maintenance on the multiprocessor system.
It is a maintenance tool that sends a request to.

In the above configuration, the processors 51 to 5N
Have the same internal configuration. For example, the processor 51 is a control C that controls the processor 51.
A PU 51a, a self-diagnosis control unit 51b that detects and checks a failure of the processor 51, and a system bus I / F unit 51d that controls the processor 51 to communicate with other processors 52 to 5N via the system bus 61. A command processing unit 51c that performs actual processing in response to a processing request from the maintenance tool 36.

The failure processing processor 14 has a system bus I / F section 14a having the same function as the system bus I / F section 51d, and when each card fails, for example, as shown in FIG. Failure information storage memory 1 for storing the failure information in which the card status is set
4b.

Further, the general-purpose communication device interface 25 controls connection and communication between the system bus I / F unit 25a having the same function as the system bus I / F unit 51d, the network 62 and the general-purpose communication device interface 25. The network I / F unit 25b, the failure information storage memory 25c having the same function and structure as the failure information storage memory 14b, and the alternative card that substitutes the function when any one of the processors 51 to 5N fails A substitute table 25d for storing substitute card information (a configuration example thereof is shown in FIG. 3), and a processing request from the maintenance tool 36 is transmitted to the network I / F unit 25.
A command check unit 25e for analyzing information such as a command and data obtained via b, and a control CPU 25f are provided. Reference numeral 21 is an external storage device for storing the replacement card information to be loaded into the replacement table 25d when the information in the replacement table 25d is stored.

The maintenance tool 36 includes a network I / F section 36a for controlling connection and communication between the network 62 and the maintenance tool 36, and a failure information storage memory 36b having the same function and configuration as the failure information storage memory 25c. Substitution table 36c having the same functional configuration as substitution table 25d, network I / F unit M31
A command check unit 36d for performing command check of the command received from the device and storing / reading data in the failure information storage memory 36b or the alternative table 36c. Note that reference numeral 31 is an operator operation unit provided with a keyboard or the like that is operated by an operator and that inputs alternative card information to the alternative table 36c.

Next, the operation will be described. 4 and 5 are flow charts showing the failure processing of the multiprocessor system having the above configuration and the flow of requests from the maintenance tool 36. The processor is capable of N connections, but here, the failure processing will be described assuming that a failure occurs in the processor 51. Incidentally, the processing from the occurrence of a failure to the disconnection of the failed card from the system bus is the same as in the prior art, so the description thereof is omitted here.

Now, when a failure such as a memory parity error or WDT occurs in the processor 51, the self-determination control unit 5
1b recognizes the failure and generates failure information (S1).
The control CPU 51a sends this failure information to the system bus I /
The F unit 51d is notified as failure information, and the system bus I /
Upon receiving this failure information, the F unit 51d sends it to the failure processing processor 14 (S2).

In the failure processor 14, the system bus I / F unit 14a receives the failure information from the system bus I / F unit 51d via the system bus 61 and stores it in the failure information storage memory 14b. The data is transmitted to the system bus I / F unit 25a of the general-purpose communication device interface 25 via 61 (S3).

In the general-purpose communication device interface 25, the command check unit 25e has a system bus I / O.
The failure information from the F section 25a is stored in the failure information storage memory 25c.
(S4) and the network I / F unit 25
Notify b. Further, the network I / F unit 25b receives the network I / F unit 36 via the network 62.
The failure information received from the command check unit 25e is communicated to a.

The network I / F unit 36a receives the failure information via the network 62, and the command check unit 3
6d is notified (S5). Then, the command check unit 36d stores the failure information in the failure information storage memory 36b (S6).

When the maintenance tool 36 sends some processing as request information to the processor 51 after the processor 51 falls into the failure state as described above and is disconnected from the system bus 61 (S7), the request information 42 is accepted. Then, the command check unit 36d refers to the failure information storage memory 36b to check whether or not the card to be communicated is in a failed state (S8).

For example, in FIG. 2, when there is a request to the processor 52, the processor 52 determines that there is a serious failure. If it is in the failure state (S9), the substitution table 3
It is checked whether the substitute card information is written in 6c (S10). For example, when the alternative table 36c is registered (S11), the command check unit 36d checks the alternative table 36c and recognizes the communicating processor (S13). For example, alternative table 36c
3 has a configuration as shown in FIG. 3, the substitute card for the processor 52 having the above-mentioned serious failure is the processor 53.

If the contents of the substitute card information are not defined like the processor 53, the operator operation unit 3
By notifying 1 of the above, for example, the operator is prompted to input the information, and the input of the operator notifies the command check unit 36d of the alternative card information, and the command check unit 36d stores the alternative card information in the alternative table 36c ( S12).

A method of designating the alternative card information is, for example,
The replacement card to the processor 52 next to the processor 51,
The processor 52 describes the next processor like the processor 53, and the command check unit 36d generates the alternative card information according to the method input by the operator or the failure information generated in the failure processor 51, and the failure processing processor 14 There is a method of sending data to a card such as. For example, the command check unit 32 refers to the substitute card information 38 in the substitute table 34 as shown in FIG. 3, and substitutes the function of the card with which the maintenance tool 36 communicates such that the substitute card of the processor 51 is the processor 52. The card to be used (hereinafter, a substitute card) is recognized and designated (S13).

The command check unit 36d is used by the network I / F unit 36a to specify the processor 5 designated as described above.
The second alternative card (processor 53 in this embodiment) is notified as the request destination processor (S14). Then, the network I / F unit 36a transmits a request to the substitute card, so that the maintenance tool 36 is connected to the processor 53 that is the substitute card and communication is possible, and the processor 53 requests the maintenance tool 36. Is processed as a command (S15).

Example 2. In the above embodiment, the replacement table 36c is provided in the maintenance tool 36, but the replacement table 25d and the failure information storage memory 25c are provided in the general-purpose communication device interface 25 as shown in FIG.
The maintenance tool 36 is, for example, the processor 51.
When the processor 51 is in the failure state when the communication is requested, the command check unit 25e in the general-purpose communication device interface 25 has the same effect even if the failure detection and the designation of the substitute card are performed.

Example 3. In the above embodiment, the processor 51
Although the failure information generated in step 1 is once sent to the general-purpose communication device interface 25 via the failure processing processor 14, the same effect can be obtained even if the failure information is sent directly from the processor 51 to the general-purpose communication device interface 25. can get.

[0028]

As described above, according to the present invention, when a maintenance tool requests a maintenance function to a processor in a multiprocessor system, even if the concerned processor is severely broken and is disconnected from the system bus. As long as another processor substitutes the function, the maintenance tool can request the substitute card for processing without the operator having to be aware of the maintenance tool, which has the effect of improving the maintainability of the system.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing a first embodiment of the present invention.

FIG. 2 is a content diagram of a failure information storage memory of the present invention.

FIG. 3 is a content diagram of an alternative table of the present invention.

FIG. 4 is a flowchart showing a failure processing flow of the present invention.

FIG. 5 is a flowchart following FIG.

FIG. 6 is a configuration diagram of a conventional example.

FIG. 7 is a flowchart showing a conventional failure processing flow.

[Explanation of symbols]

 14 Failure Processing Processor 14a System Bus I / F Section 14b Failure Information Storage Memory 21 External Storage Device 25 General-purpose Communication Device 25a System Bus I / F Section 25e Command Check Section 25b Network I / F Section 25c Failure Information Storage Memory 25e Command Check Section 25f control CPU 31 operator operation unit 36 maintenance tool 36a failure information storage memory 36b failure information storage memory 36c alternative table 36d command check unit 51a control CPU 51b self-diagnosis control unit 51c command 51d system bus I / F unit 51 to 5N processor Card 61 System bus 62 Network

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] August 19, 1992

[Procedure Amendment 1]

[Document name to be amended] Statement

[Name of item to be corrected] 0003

[Correction method] Change

[Correction content]

In the above configuration, the processors 51 to 5N
Have the same internal configuration. For example, the processor 51 includes a control CPU 51a and a processor 51.
Self-diagnosis control unit 51b for detecting and checking the failure of
It analyzes the command to perform the actual processing, the command processing section 51c for executing the processor 5 via the system bus 61
1 includes a system bus I / F unit 51d that controls the communication with the other processors 52 to 5N.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0011

[Correction method] Change

[Correction content]

[0011]

EXAMPLES Example 1. Hereinafter, Embodiment 1 of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of the present invention, and the same parts as those of FIG. 6 showing the prior art have the same reference numerals.
61 system bus, processor 51~5N is connected to the system bus, 21 an external storage device a hard disk or the like, 62 a network, 25 is an input-output and network 62 between the external storage equipment 21 A general-purpose communication device interface that performs the interface of the system, 14 is a failure processing processor that performs failure processing of the entire system, and 36 is a processor 51 to 5N, a general-purpose communication device interface 25, and a failure processing processor 14 for performing maintenance on the multiprocessor system.
It is a maintenance tool that sends a request to.

[Procedure 3]

[Document name to be corrected] Drawing

[Name of item to be corrected] Fig. 4

[Correction method] Change

[Correction content]

[Figure 4]

[Procedure amendment 4]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 5

[Correction method] Change

[Correction content]

[Figure 5]

Claims (1)

[Claims] 1. A network is connected to a maintenance tool that sends a processing request to an arbitrary processor to perform maintenance on a multiprocessor system,
And a general-purpose communication device interface connected to the multiprocessor system via the system bus,
In a multiprocessor system having a failure processing processor for performing failure processing of the entire system, in the failure processing processor, the general-purpose communication device interface and the maintenance tool, if any processor of the multiprocessor system falls into a failure, A failure information storage memory that stores failure information of the processor is provided, and a replacement table that stores alternative processor information that replaces the function of the processor that is in failure, and whether the processor with which the maintenance tool communicates are in a failed state is checked. However, in the case of a failure, a command check unit that determines a processor to send a request from the maintenance tool based on the substitution table is provided in the maintenance tool and the general-purpose communication device interface. Multiprocessor system.