JPH02109143A - Diagnostic method for duplex system - Google Patents

Abstract

(57) [Abstract] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Summary] Regarding a diagnostic method for an online real-time information processing system consisting of a plurality of duplexed devices, when a failure occurs in the duplexed system, the device itself or a path connecting the devices is detected. The purpose of the present invention is to provide a diagnosis method for a redundant system that can easily detect the location of a problem even if it occurs. It is equipped with a fault processing section that starts when a fault is detected in the active system, and the state storage means of the fault processing section stores the state information of the system configuration at that time upon detection of the fault, and the switching means saves the state information of the system configuration at that time. According to the state of the system configuration determined, all the devices involved in the occurrence of the failure are incorporated into the route to be diagnosed and configured to drive the diagnostic means.

[Industrial Application Field] The present invention relates to a diagnostic method in an online real-time information processing system consisting of a plurality of duplex devices.

In information processing systems such as electronic exchanges that perform online real-time processing, it is not allowed to interrupt the processing even momentarily, so the main equipment is duplicated, with one system performing online processing as the active system. , the other system is set as a standby system so that it can become active at any time.

When a failure occurs in the active system, the standby system is switched to the active system, and the faulty device detected online is diagnosed using a diagnostic program from the offline standby system. It may take time and effort to identify the location, and improvements are desired.

[Prior Art] FIG. 5 is an explanatory diagram of a conventional example, and FIG. 6 is a processing flow diagram of the conventional example.

In FIG. 5, the duplex system has the same configuration as the active (indicated by ACT, which means action) system and the standby (indicated by S 13 Y, which means standby) system, and main memory (indicated by MM) 01, Central controller (denoted CC) 0.1 and channel controller (CHC)
) 0 and 1 are provided. As is well known, in the case of an electronic switching system, a large number of call processing processors connected by a lotus are connected to the CC, and an input/output control device and an external storage device are connected to the CHC (not shown). .

FIG. 5A shows the failure detection state, MMICCI,C
HCl is ACT system (online), MMO, CCO,
(l(Assume that a failure occurs while Co is performing real-time processing as an SBY system. In this case, information on which device the failure occurred in is also known, and in the case of A, AC
CCI is disabled due to control signals such as request signals and data from T-system CHC1 not being given to CC1 for a predetermined period of time.
Detects that a failure has occurred in CHCI.

In this case, the conventional processing flow shown in FIG. 6 is executed. That is, when it is determined that a fault has been detected online (step 60), the faulty device is disconnected and the corresponding SBY system device is connected to the ACT system (online) (step 61). Next, a diagnostic program is executed offline for the faulty device (step 62).

This state is shown in Figure 5B, where CHCO is an ACT system,
The failed CHC1 is taken offline, connected to the MMO and CCO as shown by the - dotted chain line, and diagnosed. When there is a fault in the route between CCI and CHCl, the fault cannot be detected by the diagnostic program because the fault location is not included in the route to be diagnosed.

Next, in step 63 of FIG. 6, it is determined whether or not a failure point has been detected by the diagnosis in step 62. If it is detected, the process ends; if it is not detected, a manual is provided that describes the condition of the failure and how to deal with it. The central controller CC localizes the fault location by referring to
etc. system (step 64), and execute the diagnostic program (step 62). In the example of FIG. 5, the switching/diagnosis state of the control system is set as shown in C1. In this case, since no failure point was detected in CHCl by the diagnostic operation shown in FIG. 5B, a switching operation is performed to connect CHC1 offline via MMO and CC1 to perform diagnosis.

If this diagnosis detects that there is a fault in the path from CC1 to CHCl, it can be determined that the fault has occurred in the interface circuit between CCI and CHCl.

[Problems to be Solved by the Invention] According to the above-mentioned conventional method, when a failure occurs in the connection path between devices, the operator must perform a complex process to identify the failure point. There is a problem in that it is necessary to perform the switching operation several times, which requires time and effort.

According to the present invention, when a fault occurs in a redundant system, the location of the fault can be easily detected even if the fault occurs in the device itself or the path connecting the devices, and all points where the fault is suspected can be quickly isolated. The purpose of this paper is to present a diagnostic method for a redundant system that is capable of iiJ functions.

[Means to solve the problem 1 FIG. 1 is a diagram showing the basic configuration of the present invention.

In FIG.
1 represents a memory section, and 12 represents a connection path selection section.

In the present invention, when a fault occurs, the state at the time of the fault is saved, and the device is switched to perform diagnosis so that the saved state is incorporated into the diagnosis route.

[Operation] When a signal indicating the occurrence of a failure is input to the failure processing unit 10 of the central control unit CC which is performing online processing as shown in FIG. The status storage means 112 retrieves the contents from the management table 111 in which the system configuration is stored.
Store in.

Next, the switching means 102 switches off all the devices that were operating in the active system when the failure occurred, and brings the corresponding standby (SBY) devices online in place of those devices, so that they can be used in the active system at the time of the failure. system, and instructs the central control unit CC, which is currently offline, to perform diagnosis.

As a result, diagnosis is executed by the diagnostic means 103 on the offline side, and by the diagnostic operation after this switching, it becomes possible to detect a failure that has occurred between the devices.

[Example] Fig. 2 is a processing flow diagram of an embodiment of the present invention, Fig. 3 is a diagram showing an example of a management table and related device table, and Fig. 4 is a configuration diagram of a duplex system in which the present invention is implemented. show.

In FIG. 2, when a failure occurs (20), the state information of the system configuration at this time is saved (21). The state of this system configuration is shown in Figure 3A.

The contents of the management table (secured in a specific area in the main memory) are extracted and stored in a work area for troubleshooting. A, currently online operation is performed for device numbers 0, 1, 2, 3, 4, etc. of each central controller CC, main memory MM, channel controller CHCl input/output controller IOC, etc. 5erv
ice) at 1°, and in other conditions, i.e., O
US (disconnected state) or standby SBY (state that can immediately become ACT) is displayed as ``0''.

Next, referring to the diagnosis-related device table, the state of the related device in the saved state information is switched (22). This diagnosis-related device table is a table in which a plurality of devices that are closely related when diagnosing a device to be faulty is stored as a table, and one example is shown at the beginning of FIG. mouth,
In the table, devices related to diagnosing the central control unit CC, main memory MM, and channel control unit CHC, which are each diagnostic device, are stored as a table. For example, it is impossible to diagnose only the central control unit CC. The main memory MM and channel control device CHC operate in conjunction with each other.

After the above switching is performed, the standby system executes diagnosis (23). At this time, other ACT-based devices execute online processing.

FIG. 4 is a configuration diagram of a duplex system in which the present invention is implemented. When MMO, CCO, CHCO and l0C (input/output control unit) 0 in the figure are on standby as SBY system (OUS state), and MMI, CCI, CHCl and l0CI are executing online processing as ACT system, each prefecture Selection circuits 40 to 43 are connected to select the bus on the solid line side, respectively.

For example, if a failure occurs in CHC1 in this state, the process shown in FIG.
” and other devices are “0”) is saved, and the device that was in the SBY system at the time of the failure becomes the ACT system, and the device that was in the ACT system becomes the SBY system.In this state, the selection circuit 4
0 to 43 are controlled to select the solid line side.

As a result, MMl, CC1, and CHCI, which have become SBY systems, are connected, and the diagnosis program is executed from CC1 and MMI to diagnose CHC1. If a fault is detected by this diagnosis, it is determined that the fault is located in the circuit (connection line or interface circuit) between CHC1 and CCI.

[Effects of the Invention] According to the present invention, it is possible to increase the rate of detecting a failure location in a short time without requiring manual effort, and it is possible to improve the reliability of the system.

A system configuration diagram, FIG. 5 is an explanatory diagram of a conventional example, and FIG. 6 is a processing flow diagram of the conventional example.

10: Failure processing unit 101: State storage means 102, switching means 103: Diagnostic means 11: Memory part 12: Connection path selection section Patent applicant: Fujitsu Limited Sub-agent Patent Attorney Hosaka Airi No.

[Brief explanation of drawings]

Fig. 1 is a diagram showing the basic configuration of the present invention, Fig. 2 is a processing flow diagram of an embodiment of the present invention, Fig. 3 is a diagram showing the features of a table, and Fig. 4 is a duplication diagram in which the present invention is implemented. 1, 1 31 ・i. Example of management table FIG. 2 is a processing flow diagram of the embodiment. FIG. 3 is a diagram showing an example of a table. BY ACT A. Fault detection state offline ACT B, faulty device switching/diagnosis state offline ACT

Claims (1)

[Scope of Claims] An information processing system that performs online real-time processing consisting of a plurality of duplex devices of an active system and a standby system, comprising a fault processing unit (10) that is activated when a fault is detected in the active system. The state storage means (101) of the fault processing unit (10) stores the state information of the system configuration at that time upon detection of a fault, and the switching means (102) changes the system configuration state to the state saved by the state storage means. A method for diagnosing a redundant system, characterized in that all devices involved in the occurrence of a fault are incorporated into a route to be diagnosed and a diagnostic means (103) is driven accordingly.