JP3156987B2 - Bus failure diagnosis method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for diagnosing a bus failure in a system in which a plurality of devices are connected to a common bus as an information transfer route when an abnormality is detected during communication via the bus.

In particular, in systems where high reliability and maintainability are desired, such as switching systems and information processing systems, system abnormalities can be assured automatically without maintenance personnel in response to this type of communication error on the bus. It is expected that failure diagnosis can be specified in a short time.

[0003]

2. Description of the Related Art Generally, a system requiring high reliability such as a switching system employs a redundant configuration in which each device is duplicated so that processing of the system can be continued even if a device fails. Further, from the viewpoints of economy and expandability of the device, it is general to adopt a bus connection method as a connection configuration of the device.

FIG. 2 is an example of a system configuration diagram in accordance with the above-described concept. Hereinafter, the prior art will be specifically described with reference to the configuration example of FIG. In this system configuration example, the duplicated device group forms two symmetric systems (0 system and 1 system),
Central processing unit 010, peripheral device 03
The central bus 100 is connected to the central processing unit 110 and the peripheral devices 131 to 13n. The 0-system bus 000 and the 1-system bus 100 are duplicated cross devices 0
20 and 120 and an intersection bus 200. Here, the communication route between the two devices is the only intersection.

[0005] When a fixed failure occurs in the system, it is necessary to identify the device containing the failure and disconnect it from the system in order to restart the processing normally as a system. If each device detects an abnormality in the course of its internal processing, the software determines that the notification source device has failed by the failure notification from that device, and can immediately disconnect the failure processing. However, when an abnormality is detected during communication from a certain device to another device via a bus, a plurality of devices involved in the communication may be considered as a possible cause of a failure point.

FIG. 3 shows a non-response (abnormal bus communication) when software (0-system central processing unit 010) accesses the 1-system peripheral device 131 in the configuration example shown in FIG.
FIG. 9 is a diagram showing a suspected range when the condition becomes. In this case, the suspected range is all devices on the communication path. 0 system central processing unit 010 0 system bus 000 0 system crossing device 020 crossing bus 200 1 system crossing device 120 1 system bus 100 It is considered to be one of the system 1 communication destination peripheral devices 131, and a method for diagnosing a fault location (diagnosis)
Is an important issue.

Conventionally, there is a method called a rotation method as a method of isolating a faulty device when this type of bus communication is abnormal. The rotation method is a method in which the processing is restarted by simply switching the redundant devices that may have a failure one by one in a predetermined order. If the faulty device is disconnected during this switching process,
The processing after the restart is executed normally. If the faulty device has not been disconnected, the abnormality is detected again, and the next device is switched according to the procedure and the process is restarted again. By repeating such device disconnection and process resumption, the failed device can be eventually eliminated from the system.

FIG. 4 is a diagram for explaining a conventional rotation method. The 0-system central processing unit 0 shown in FIG.
An example of a system configuration pattern and a change order of a rotation to be executed when a bus communication from 10 to the 1-system peripheral device 131 becomes unresponsive (bus communication abnormality) is shown.

In the system configuration pattern shown in FIG.
Indicates a state of being incorporated into the system in the corresponding pattern. In addition to the peripheral devices 031 and 131, ○ indicates a state in which all devices are incorporated in the corresponding system, and △ indicates a state in which some devices are incorporated in the corresponding system. The number under the device type name indicates the device number in FIG. * Indicates 0 (0 series) or 1
(1 system) is shown.

In the case of the system configuration change order shown in FIG. 4B, for example, if the failure point of the bus communication abnormality shown in FIG. It is disconnected and normal processing can be resumed.

However, at this stage, the 0-system bus 000,
0 system crossing device 020, crossing bus 200, 1 system crossing device 1
Since it is not possible to identify which of the devices 20 is faulty, after the normal processing is restarted, the maintenance person executes a normality test on each of the simulated devices from the terminal and specifies the faulty device.

[0012]

In the above-mentioned rotation method, there is a problem that it takes a long time until the faulty device is finally separated because the system switching process of the device is involved. The time increases exponentially as the number of devices to be simulated increases. For example, assuming that the average required time from switching the system of a device to restarting the process and detecting an abnormality again is t, and the number of devices to be simulated is n per device, the required time until the faulty device is separated is the worst. It can be t × (2 ⁿ −1).

Further, by the time a normal system is finally constructed,
There is a possibility that a normal device that does not need to be system-switched may be switched, and there is a problem that erroneous processing (temporary interruption of processing) is caused by the system switching.

Furthermore, the purpose of the rotation method is merely to isolate a faulty device (restart normal processing), and it may not be possible to specify a final faulty device as in the example shown in FIG. In this case, it is necessary for the maintenance person to specify the failure location, which increases the work of the maintenance person.

The present invention has been made in view of the above circumstances, and has been made to speed up failure diagnosis when a bus communication abnormality is detected, to avoid erroneous processing due to normal system switching, and to reduce maintenance work. It is an object of the present invention to provide a bus failure diagnosis method which can be performed.

[0016]

In order to achieve the above object, a bus failure diagnosis system according to the present invention comprises a plurality of devices including a central processing unit which are connected to one common bus or a plurality of common buses via a relay device. In a system comprising a single or a plurality of bus systems connected to a, a failure diagnosis method when an abnormality is detected during communication via the bus between a central processing unit and a device other than the central processing unit, The central processing unit, a program for testing the normality of the internal circuit of each device, a program for testing the normality of the communication function via the bus from the central processing unit to each device, a diagnostic scenario, A program for interpreting and executing the diagnostic scenario, wherein the diagnostic scenario first executes a diagnosis of a bus system including a central processing unit, and a plurality of bus systems connected via a relay device. In the case where the bus system including the central processing unit is diagnosed, the bus system is diagnosed in order from the central processing unit to a bus system far away from the central processing unit. Is executed from a device close to the central processing unit, and the diagnostic scenario for diagnosing a bus system to which three or more devices including the central processing unit or the relay device are connected is an arbitrary one other than the central processing unit or the relay device. It is configured to test the normality of the communication function via the bus for the two devices, and to test the internal circuit of the central processing unit or the relay device when the test results for both devices are both abnormal. The presence or absence of a bus failure is determined from the result of access to three devices connected to the same bus, and the central processing unit detects an abnormality during communication via the bus. As a trigger, before disconnecting from the system any device, it is characterized in that identifying the defective unit by executing the diagnosis scenario.

Further, according to the present invention, in the bus failure diagnosis method, the diagnosis scenario for diagnosing a bus system to which two devices including a central processing unit or a relay device are connected may include a diagnostic system other than the central processing device or the relay device. The device is configured to test the normality of the communication function via the bus with respect to the device, and to test the internal circuit of the central processing unit or the relay device when the test result is abnormal. Things.

[0018]

[0019]

[0020]

According to the present invention, in a system in which a plurality of devices are connected to a common bus as their information transfer routes, when an abnormality is detected during communication via the bus, the failure location is determined. This is a diagnostic method that enables autonomous and reliable identification in a short time.

The central processing unit includes a program for testing the normality of each device internal circuit, a program for testing the normality of communication from the central processing unit to each device via the bus,
A diagnostic scenario configured by combining these test programs; and a program for interpreting and executing the diagnostic scenario. When the central processing unit detects an abnormality during communication via the bus, any device is operated by the operating system. Before disconnecting from the device, a test program is executed on the suspected device on the communication path at the time of detecting the abnormality according to the diagnostic scenario, and the faulty device is identified from the test result.

[0022]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a flowchart of a failure point specifying process in the present embodiment. This flowchart is shown in FIG.
In the example of the system configuration shown in FIG. 3, as shown in FIG. 3, there is shown an embodiment in which the communication from the 0-system central processing unit 010 to the 1-system peripheral device 131 has no response (bus communication abnormality).

That is, the failure location in this case is one of the following devices as described above. -System 0 central processing unit 010-System 0 bus 000-System 0 crossing device 020-Crossing bus 200-System 1 crossing device 120-System 1 bus 100-System 1 communication destination peripheral device 131 In this embodiment, the software diagnoses a failure. Perform the processing,
The procedure is as follows. In the following description,
The normality test of communication via the bus from the central processing unit to the corresponding device is referred to as "test A", and the normality test of the corresponding device internal circuit is referred to as "test B".

The test A can be executed by, for example, writing, reading, and collating a test pattern in a control register in a corresponding device. The test B may be a method of starting an autonomous test provided in each device or a method of executing the test under control of a higher-level program from a central processing unit. When the autonomous test provided in each device is started, it may be started on a route different from the bus.

When the software detects a bus communication abnormality, the test first executes [Test 10] Test A on the 0-system crossing device 020. If the result of this [Test 10] is NG, [Test 11] any device other than the central processing unit 010 and the tested intersection device 020 among the devices connected to the 0-system bus (in this embodiment, The test A is performed on the device 031). If the result of [Test 11] is NG, the simulated fault connected to the 0-system bus is 0-system bus 0
Since there is a possibility of the central processing unit 010 of system 00 and system 0, [Test 12] 0
The test B is executed for the system central processing unit 010. If the result of this [Test 12] is NG, the system 0 central processing unit 0
10 failure (13), if the result is OK, 0-system bus 00
It can be determined that the failure is 0 (14). If the result of [Test 11] is OK, since the 0-system bus 000 is normal, it can be determined that the 0-system crossing device 020 has failed (22). [Test 1
If the result of [0] is OK, then [Test 20] test A is performed on the 1-system crossover device 120. This [Test 2
0] is NG, the system 0 crossing device 020, the system 1 crossing device 120 and the crossing bus 20
[Test 2]
1] Test B is performed on the 0-system crossing device 020. If the result of this [Test 21] is NG, the 0-system crossing device 02
0 failure (22), if the result is OK, 1-system crossing device 1
20 or the failure of the crossing bus 200 (23). If the result of [Test 20] is OK, then [Test 3]
0] Test A is performed on the first system peripheral device 131 which is the communication destination device when the bus communication abnormality is detected. This [Test 30]
If the result of the test is NG, then [Test 31] the 1-system bus 10
The test A is executed for an arbitrary device other than the intersection device 120 and the peripheral device 131 (in this embodiment, the first system central processing device 110) among the devices connected to 0.

When the result of this [Test 31] is NG, the system 1 bus 100 and the system 1
Since test 0 is considered, [Test 32] test B is performed on the system 1
Execute If the result of this [Test 32] is NG, 1
It can be determined that the failure of the system crossing device 120 (33), and if it is OK, the failure of the system 1 bus 100 (34). [Test 3
If the result of [1] is OK, since both the system 1 bus 100 and the system 1 intersection device 120 are normal, it can be determined that the system 1 peripheral device 131 has failed (35). If the result of [Test 30] is OK, all devices are normal (no fixed failure device), and it is determined that the detected failure is intermittent and the operation is continued (40). .

In the description of the above embodiment, the failure diagnosis processing is executed by an online program of software, but it may be realized by firmware. Further, it may be realized by an offline program, and the same effect is obtained.

As described above, when a bus communication error is detected, the system is autonomously operated according to a predetermined procedure before disconnecting the device from the system. By performing tests sequentially on devices and identifying faulty devices from the test results,
(A) It is possible to avoid erroneous processing accompanying system switching of a normal device in the conventional rotation method. (B) The time required for system switching of the device is unnecessary, so that the time required for identifying a failure location can be greatly reduced. c) The fault location can be specified for each device autonomously by the system.

In terms of speeding up the failure location identification, the failure location identification procedure of the present invention includes (a) executing a test access to the simulated device from a device close to the central processing unit, and (b) performing the same access. It is characterized in that the presence or absence of a bus failure is determined from the result of access to three devices connected to the bus. This is a procedure that makes it possible to detect the failure of the bus itself and to minimize the average required number of tests to be performed until the failure location is finally identified. The required time is shortened.

[0030]

As described above, according to the present invention, it is possible to speed up fault diagnosis when a bus communication error is detected, avoid erroneous processing due to normal system switching, and reduce maintenance work. A bus fault diagnosis method can be provided.

[Brief description of the drawings]

FIG. 1 is a flowchart of a failure diagnosis process according to an embodiment of the present invention.

FIG. 2 is a configuration explanatory diagram showing a system configuration example in a conventional example and an embodiment of the present invention.

FIG. 3 is a configuration explanatory diagram showing an example of a simulated range when a response waiting timeout during bus communication is detected in the conventional example and the embodiment of the present invention.

FIG. 4 is an explanatory diagram showing an example of a system configuration pattern and a system configuration change order of a conventional rotation method.

[Explanation of symbols]

10, 11, 12, 20, 21, 30, 31, 32... Tests for each device in the failure diagnosis processing, 13, 14, 2
2, 23, 33, 34, 35, 40: failure location determined from the test result, 000: 0 system bus, 100: 1 system bus, 200: crossing bus, 010: 0 system central processing unit;
110 ... 1 system central processing unit, 020 ... 0 system crossing device, 1
20 ... 1 system crossing device, 031-03n ... 0 system peripheral device,
131 to 13n: 1-system peripheral device.

Continuation of front page (56) References JP-A-4-291850 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04L 12/28 H04L 29/02 H04L 29/14

Claims (2)

(57) [Claims] 1. A system comprising a single or a plurality of bus systems in which a plurality of devices including a central processing unit are connected to one common bus or a plurality of common buses via a relay device. A failure diagnosis method when an abnormality is detected during communication via the bus with a device other than the central processing unit, wherein the central processing unit tests a normality of an internal circuit of each of the devices. And a program for testing the normality of a communication function via the bus from the central processing unit to each device, a diagnostic scenario, and a program for interpreting and executing the diagnostic scenario, wherein the diagnostic scenario comprises: When a bus system including the central processing unit is diagnosed, and when there are a plurality of bus systems connected via the relay device, the bus system including the central processing unit is diagnosed, The bus system is configured to diagnose the bus system in order from the central processing unit to a distant bus system, and a test access to the device to be simulated is executed from a device close to the central processing unit. The diagnostic scenario for diagnosing a bus system to which one or more devices are connected tests the normality of the communication function via the bus for any two devices other than the central processing unit or the relay device, and If the test results for the devices are both abnormal, the internal circuit of the central processing unit or the relay device is tested, and the presence or absence of a bus failure is determined based on the result of access to three devices connected to the same bus. Before disconnecting any device from the system upon detecting that the central processing unit has detected an abnormality during communication via the bus, the diagnosis is performed. A bus failure diagnosis method for identifying a failed device by executing a disconnection scenario. 2. The diagnostic scenario for diagnosing a bus system to which two devices including a central processing unit or a relay device are connected includes a communication function via a bus for a device other than the central processing unit or the relay device. 2. The bus failure diagnosis method according to claim 1, wherein a normality is tested, and an internal circuit of the central processing unit or the relay device is tested when the test result is abnormal.