JP2680910B2 - Failure reporting method - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a failure reporting system, and more particularly to a system for reporting a failure occurring in an information processing apparatus including a plurality of modules to a failure processing apparatus.

2. Description of the Related Art In recent years, demands for high reliability or availability for information processing devices have increased, and the importance of fault handling has increased more and more.

Conventionally, when a failure occurs in an information processing apparatus, the information processing apparatus is stopped in the state where the failure has occurred, the failure processing apparatus is notified of the occurrence of the failure, and the information processing apparatus in which the failure has occurred Processing such as collecting a log dump of was being performed.

The information processing apparatus described above is composed of a plurality of modules such as LSI (Large Scale Integrated Circuit), and has been speeded up in accordance with recent technological progress and its machine cycle has been shortened. Therefore, the following method is used to stop the entire device when a failure occurs.

That is, as shown in FIG. 2, the failure detected by each module 5-i (i = 1, 2, ..., N) constituting the information processing apparatus 5 is reported to other modules, and each module is notified. The device stop flags 53-i respectively provided in 5-i are set at the same time, and the respective module 5-i are stopped simultaneously by the device stop flag 53-i.

Here, each module 5-i sets the module representative error indicator flag 51-i collectively for the failures detected in its own module, and outputs the output of the module representative error indicator flag 51-i to the module representative error reporting line 111-i. They are reported to the device stop condition generating unit 52-i in each module 5-i via i.

In the device stop condition generation unit 52-i, the failure reports input from the module representative error indicator flags 51-i via the module representative error report lines 111-i are compiled, and the device stop flag 53-i is generated according to the report contents. i is set.

The module representative error indicator flag 51-i
And, the device stop flag 53-i, once set, is kept in that state until it is released.

Since each module 5-i distributes the output of the device stop flag 53-i to its own module using the device stop signal line 112-i, each module 5-i is stopped at the same time.

Further, by reporting the output of the device stop flag 53-2 of the module 5-2 to the failure processing device 3 using the device stop signal line 112-2, it is possible to detect that a failure has occurred in the information processing device 5. The device 3 is notified.

In such a conventional fault reporting method, the fault processing device 3
When the device stop flag 53-2 of the module 5-2 that notifies the information processing device 5 of a failure has failed, there is a drawback that the failure of the information processing device 5 is not reported to the failure processing device 3.

To deal with this drawback, the outputs of the device stop flags 53-i of all the modules 5-i configuring the information processing device 5 may be sent to the fault processing device 3. In that case, The number of input / output signal lines between the device 5 and the fault processing device 3 will increase.

Also, the device stop flag 53- that should be set at the same time
Even if one of i is illegally set, it cannot be detected.

In order to deal with this drawback, the outputs of the device stop flags 53-i of the respective modules 5-i may be compared and checked with each other, and the result may be reported to the fault processing device 3. The number of input / output signal lines is increased.

SUMMARY OF THE INVENTION The present invention has been made to eliminate the above-mentioned drawbacks of the conventional ones, and can reliably report a failure occurring in an information processing apparatus to a failure processing apparatus, and set an apparatus stop flag illegally. It is an object of the present invention to provide a failure reporting method capable of reliably detecting that a failure has occurred.

According to the failure reporting method of the present invention, failure information indicating that a failure has been detected in at least one of the self module and another module is held, and the operation of the self module is stopped according to the failure information. A failure reporting method for an information processing apparatus, which comprises a plurality of modules each having a stop flag common to each module, wherein a first stop flag holding means for holding the stop flag in each of the plurality of modules, Second stop flag holding means for holding the stop puffer from an adjacent module; and the first stop flag holding means.
Or a sending means for sending the stop flag held in the second stop flag holding means to the second stop flag holding means of another adjacent module, and provided in one of the plurality of modules. In addition, it is characterized in that the sending means reports the failure of the own apparatus to the failure processing apparatus.

Next, an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of one embodiment of the present invention. In the figure, in the embodiment of the present invention, when a failure occurs in the information processing apparatus 1, until the apparatus stop flag 21-i (i = 1, 2, ..., N) is set due to the occurrence of the failure. The operation is similar to the operation until the conventional apparatus stop flag 53-i shown in FIG. 2 is set.

It should be noted that the module representative error indicator flag 51-i and the device stop condition generating unit 52-i required for that purpose.
The parts corresponding to are not shown.

The stop status invalid flag 22-i indicates the adjacent module 2-
failure report signal line 101-i from the stopped state transmitter 23-i of i
Set in response to a problem report reported via.

That is, the failure status signal line 101 is output from the failure status transmitting unit 23-2 of the module 2-2 to the failure status flag 22-1.
-A failure report is reported via -2 and the stop status is invalid.
A failure report is reported to the module 22-2 from a stopped state transmitter 23-3 (not shown) of the module 2-3 (not shown) via the failure report signal line 101-3. It should be noted that a failure report is reported to the stopped status incorrect flag 22-n from the stopped status transmission section 23-1 of the module 2-1 via the failure report signal line 101-1.

Further, once set, the device stop flag 21-i and the stop state invalid flag 22-i are kept in that state until they are released.

In the stop state transmission unit 23-i, the NAND of the output of the device stop flag 21-i and the output of the stop state incorrect flag 22-i is taken and the result is adjacent via the fault report signal line 101-i. Stop status invalid flag 22-i of module 2-i
And the failure report from the stopped status transmitter 23-i of the adjacent module 2-i is held in the stopped status incorrect flag 22-i in its own module according to the result. become.

Incidentally, the failure processing apparatus 3 is notified of a failure of the information processing apparatus 1 from the stop state transmission unit 23-2 of the module 2-2 via the failure report signal line 101-2.

The operation of the embodiment of the present invention will be described below with reference to FIG.

When no failure has occurred in the information processing device 1, it is assumed that the device stop flag 21-i and the stop state incorrect flag 22-i of the module 2-i are each cleared to "0".

First, the device stop flag 21 of all modules 2-i
When -i is set normally and at the same time, the device stop flag 21-i is set so that all the stop state invalid flags 22-i keep the state cleared to "0".

For the failure processing device 3, the device stop flag 21-i
Is set is reported from the stopped state transmitter 23-2 of the module 2-2 via the fault report signal line 101-2.

Further, the device stop flag 21-2 of the module 2-2 is not set due to a failure, and the other modules 2-1 and 2-3 to
When the device stop flags 21-1, 21-3 to 21-n of 2-n are normally and simultaneously set, a failure report is sent from the stop state transmission unit 23-3 of the module 2-3 adjacent to the module 2-2. The fact that the device stop flag 21-3 (not shown) is set is transmitted via the signal line 101-3, and the module 2
-2 stop state incorrect flag 22-2 is set.

As a result, the stopped state transmitter 23- of the module 2-2
2 through the fault report signal line 101-2, the fault processing device 3
Is reported that a failure has occurred in the information processing device 1.

However, in this case, except for the device stop flag 21-2 of the module 2-2, the other modules 2-1 and 2-3-2
-N device stop flags 21-1, 21-3 to 21-n are set, and the stop state invalid flag 22-i of the module 2-i is set.
Since all are set, module 2-2
It is possible to know that the device stop flag 21-2 of 1 is out of order.

Furthermore, when the device stop flag 21-i is illegally set in any one of the modules 2-i, the stop state illegal flag 22-i of the adjacent module 2-i is sequentially set, and the module 2-2 Stop status invalid flag 22
-2 is set, the failure state transmission unit 23-2 of the module 2-2 reports the occurrence of a failure in the information processing apparatus 1 to the failure processing apparatus 3 via the failure report signal line 101-2. .

In this way, the stop status invalid flag of module 2-i
A failure report from the module 2-i adjacent to the module 22-i is held, and the value of the stop status invalid flag 22-i and the device stop flag 21 for stopping the operation of the own module 2-i are held.
-The value of -i is used to generate a fault report of its own module 2-i in the stopped state transmission unit 23-i, and another adjacent module 2
-I to send these multiple modules
By letting one of the i's stop state transmission units 23-i report a failure of the own apparatus to the failure processing apparatus 3, it is possible to add an amount of hardware and an input / output signal line to the information processing apparatus. It is possible to reliably report the failure that occurred in No. 1 to the failure processing device 3 and to reliably detect that the device stop flag 21-i has been set illegally.

In the embodiment of the present invention, the failure status flag 22-i and the stopped status transmitter 23-i of each module 2-i are connected in a ring by the failure report signal line 101-i to report the occurrence of a failure. However, they may be connected in a tree shape, and the invention is not limited to this.

As described above, according to the present invention, failure information indicating that a failure is detected in at least one of the own module and other modules is held, and the operation of the own module is performed according to the failure information. The failure information held in the stop flag that causes the module to stop and the stop status report from the adjacent module generate the stop status report of the own module and send it to the other adjacent module. By reporting the fault of its own device to the fault processing device from one of the above, it is possible to reliably report the fault that occurred in the information processing device to the fault processing device, and the stop flag was set incorrectly. Has the effect that it can be reliably detected.

[Brief description of the drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, and FIG. 2 is a block diagram showing the configuration of a conventional example. Description of symbols of main parts 1 ... Information processing device 2-1 to 2-n ... Module 3 ... Fault processing device 21-1 to 21-n ... Device stop flag 22-1 to 22-n ... Stop Invalid status flag 23-1 to 23-n ... Stop status transmission unit

Claims (1)

(57) [Claims] 1. A stop common to each module that holds failure information indicating that a failure has been detected in at least one of the own module and another module, and stops the operation of the own module in accordance with the failure information. A failure reporting method for an information processing apparatus, which comprises a plurality of modules each having a flag, comprising: first stop flag holding means for holding the stop flag in each of the plurality of modules;
Second stop flag holding means for holding the stop flag from an adjacent module, and the second stop flag of another adjacent module for the stop flag held by the first or second stop flag holding means. A failure reporting method characterized in that a sending means for sending to a holding means is provided, and the sending means provided in one of the plurality of modules reports a failure of the device itself to a failure processing apparatus. .