JPH02140036A - System constitution decision system at the time of restarting operation - Google Patents

Abstract

PURPOSE:To detect a device having a fault before the restart and to realize system restart quickly and suitably by executing quick restart processing without integrating the device having a fault into the system at restart. CONSTITUTION:When a chance of restarting is given by a fault detection circuit BSE or the like, the system constitution of an active ACT system before restart stored in a RAM (not shown) or the like and the restarting operation is started. Then a microprogram is started and devices entirely opposite to a central controller CC or the like stored as the system constitution device before restart are used for the system rasing. Then the initial setting and clock source selection for the raising CC or the like are implemented. When the necessity of memory MM changeover is discriminated by the discrimination of fault information, a proper memory route is set. When the changeover is required, the initial setting of the MM used newly is implemented and the CC or the like is connected according to the decided system constitution.

Description

[Detailed Description of the Invention] [Summary] This invention relates to a system configuration determination method for restarting a redundant electronic exchange system to determine what kind of system configuration to take after a failure occurs. The purpose of this system is to quickly execute restart processing without incorporating the device into the restart system, and the central control unit, channel control unit, and storage device are
In the restart operation of the redundant electronic exchange system, the configuration of the active system before restart among the redundant equipment is stored, the restart operation of the system is started, and the configuration of the active system before restart is stored. Configure the system to determine the system configuration upon restart without incorporating it.

[Industrial application field]

The present invention relates to a system configuration determination method for restarting a redundant computer-controlled system, and more specifically, to determine what kind of system configuration to take after a failure occurs in a redundant electronic switching system. This invention relates to a system configuration determination method when restarting a storage device.

[Conventional technology]

Central control unit 1 (CC) of the computer control system,
Channel control unit (CHC) and memory (MM)
Duplicate systems are currently used in various fields, in which the system is duplicated into an act system (active system) and a standby system (non-active system), and a switch is made to the standby system in the event of a failure in the act system.

FIG. 5 shows a conventional method, that is, a rotation method, for determining the system configuration at the time of restart in a restart operation for restarting the system after a failure occurs in a duplex electronic exchange system.

In the figure, there are two systems for the central control unit (CC), channel control unit (CHC), and memory (MM), respectively, O system (#0) and 1 system (#1).
While C is in one set, MM is independent. That is, for example, #QCC/#(IcHcI is #
(Can be connected to either MM3 or 91MM4.

In FIG. 5, the system configuration is defined by state values, and there are four configuration methods with states 0 to 3. In state O, #OCC/#0CI (C
1 and # I MM 4, but #OCC/# in state 1
0CHC1 and 90MM3 are #ICC/ in state 2.
#ICHC2 and 91MM4 are O or #I in state 3
CC/#ICHC2 and #OMM3 are connected.

In the conventional method shown in Figure 5, restarting is attempted with a state value that is randomly determined during restart processing, and if restarting is not possible, the state is changed to the next value after the randomly determined value, as the name of rotation indicates. The method used was to update the state according to the rotation and attempt to restart, and if restart still failed, the state was updated and restarted again. In the figure, °r! is indicated by an arrow between blocks showing the system configuration for each state value. , MA'
is an abbreviation for emergency and indicates system restart processing.

FIG. 6 shows a flowchart 1 of a conventional system configuration determination method at the time of system restart operation. In the same figure, the trigger for starting the restart operation of the system is the failure detection circuit of S5a (BSE
) or the S5b, 35C central control unit (
CC) side. S5b is #OC
C, S5c is activation from #ICC. Activation from the CC side is caused by an overflow of the timer provided to detect program runaway, or by the emergency button on the console (CNS).
takes the form of

When the restart operation is started in S6, control from the next step 87 is performed by the emergency start hardware. In S7, the randomly set state values in the rotation method of FIG. 5 are updated, and then in S8, the initial settings of the central control unit (CC) and channel control unit (CI-IC) determined by the state values are performed. And the clock? R selection is made. The contents of the initial setting are as shown in the same figure, and differ depending on whether a startup CC or a non-startup CC is selected.

Among these settings, "8CT Fl" indicates whether the flip-flop is currently in use or not, and is either active (ACT) or standby (SBY).

"SP route" indicates a network, and the startup CC is connected to the same network as the value before EMA startup.
The 'SYD display' is a system down display and is turned off because it is when the restart operation is started.

'l5OI' indicates emergency activation and 'CL
K 5ELF' indicates that the clock supply source is self-system. Also, parity control (ParityCTL)
), error correction circuit controller call (ECCCTL), and testing test flag are all off for raised and non-raised CCs.

When the initial setting is completed, the microprogram (μmP) is started for the start-up CC at step 39a, and the non-startup CC is dynamically stopped at step S9b, that is, the microprogram is also stopped.

Thereafter, restart processing control is performed by firmware on the start-up CC side. Initialize memory (MM) with S10,
That is, the memory route is set, the memory of the transfer area is cleared, and the Sll connects the routes of the CC3MM, CHC, and SP (network) that constitute the system according to the above-mentioned state value. Thereafter, in step 312, the failure information that caused restart startup is saved, and in step S13, initial settings for software startup are performed. At this point, control is passed from the firmware to the software, and in step 314, the software starts controlling from the command at address 1000, for example.

[Problem to be solved by the invention]

The problems of the rotation method as a conventional system configuration determination method described in FIG. 5 will be explained with reference to FIG.

As described at 87 in FIG. 6, after the restart process is started, the state value, for example, the state counter value, is updated,
The system configuration at the time of restart processing is determined. In FIG. 7, 15 indicates a state where a failure has occurred in the memory #IMM4 and restart processing has been started. This corresponds to state 2 in Figure 5, but on the other hand, since the value of the state counter that is updated each time the restart process is started is unrelated to the system configuration at the time of startup, which state is the restart process in? It is undefined whether this is done by value.

Therefore, if restart processing is executed in state 0 as shown in 16 in FIG. 7, for example, restarting will not be possible due to the failure of 91MM4. And again in FIG. 6, for example, S5b
The flow from 1 to 3 is repeated, and the stay 1 value is set to 1 in S7, and the restart is completed by the configuration 17 in FIG. After the restart is completed, #I
A failure of MM 4 is detected.

As described above, in the conventional system shown in FIG. 5, in the system restart operation of a duplex electronic exchange system, the configuration at the time of system restart is determined according to a certain rotation, which is the cause of the system restart operation. There are cases in which a faulty device is incorporated into the system at the time of restart, and cases in which unnecessary switching is performed, resulting in a problem that the restart process takes a long time.

An object of the present invention is to promptly execute restart processing without incorporating a failed device into a system at the time of restart.

[Means and operations for solving the problems] A functional block diagram of the present invention is shown in FIG. Same figure! 1ml, (b) is a functional block diagram of the first and second inventions, respectively. In the same figure (al), when restarting an electronic switching system in which a central control unit, a channel control unit, and a storage device are duplicated, 20 indicates the current system configuration before restart, for example, a central control unit, a channel control device, and a storage device. In order to store the combination, the combination is stored in, for example, a random access memory (RAM) not shown.

The second engineer then initiates a system restart operation.

22 reads the system configuration before restart stored in 20 from, for example, RAM, and determines the system configuration without incorporating the devices that were installed in the current system before restarting the system.

In FIG. 1(b) showing the functions of the second invention, 21 has the same function as in FIG. 1fa). 23 is 21
After restarting at 312 in FIG. 6, for example, the failure information saved at 312 in FIG. 6 is utilized to determine the device in which the failure has occurred. Here, the failure information corresponds to, for example, a signal indicating an abnormality in hardware stored in a group of flip-flops (FF) in the system.

Thereafter, 24 uses the determination result of 23 to determine the system configuration at the time of restart processing without incorporating the failed device.

By executing the restart processing using the system configuration determined as described above, accurate restart processing is performed.

〔Example〕

A flowchart of a first embodiment of the system configuration determination method according to the present invention is shown in FIG. 525a, 525b, similar to S5a, S5b, and S5c in the conventional example in FIG.
, 525c failure detection circuit (BSE), #0CC1 or #ICC, when an opportunity to start the restart operation is given, in S26, the system configuration of the active (ACT) system before restart, that is, CC, C)ic , MM are stored in a random access memory (RAM) (not shown) or the like, and a restart operation is started in S27.

After that, the microprogram is started in S28, and S29
The system configuration device before restart is the opposite of the CC and CHC stored in RAM etc. For example, in FIG. 5, #OCC/#0CI (if C1 is stored, #
I CC/#I C1-IC2 are used for startup.

At step 330, the initial setting of the start-up CC and CI-(C) and selection of one cock source are performed.

Next, at 331, a determination of fault information is made to determine if memory switching is required. Here, the failure information refers to the FF (
For example, when a system error occurs due to clock interruption or power interruption in the CC, or a parity focus error in the memory, the FF corresponding to the error is turned on. Therefore, failures such as CHC and MM are determined by checking the status of each FF.

When it is determined whether or not MM switching is necessary based on the failure information '1-1, an appropriate memory loop 1- is set in S32, and if switching is necessary, the newly used MM is set in S32. The first period setting is performed, and in S34, the CC, CHC, MM, and SP (network) are connected according to the determined system configuration. If MM switching is not required, select S33.
334 is executed after the process of 832 without going through. Thereafter, control is passed from the firmware to the software in the same manner as in 313 and S14 in FIG. 6 of the conventional example.

Figure 3 shows the flow of the second embodiment of the system configuration determination method.
It is a chart. 525a, 525b in FIG.
535a, 535b, 5 exactly as in 525c
When the timing for starting the restart operation is given at 35c, 33G
At this point, the ACT FF for the non-active CC before restarting is turned on. That is, among the initial M setting contents explained in 88 of FIG.
TFF value is “ACT”, AC for active CC
The value of TFF is set to SBY'. As a result, for CC and CIf C, 326 and 329 in Figure 2
This results in substantially the same effect as storing the active system configuration in .

Thereafter, the restart operation is started in S37, and the microprogram is started in 33B. In Figure 2, after executing 328, S
The CC in the system opposite to the CC stored in step 29 is set as the startup CC, but in FIG. 3, the non-active system CC before restart has already been selected as the startup CC in step 836. Therefore, the processing corresponding to 329 is unnecessary. Thereafter, the processes from S39 onwards are executed in the same way as the processes from 330 onwards in FIG.

FIG. 4 shows an embodiment of system restart processing execution according to the present invention. Assume that a failure occurs in the #IMM 4 at 45 and restart processing is started, similar to 15 of FIG. In step 46, the non-active CC2 before restart, ie, #OCC, is selected as the startup CC, and the fault in #IMM4 is recognized in step 47 by determining the fault information. In other words, a failure was detected before restart.

After that, # OMM 3 is selected as the start-up MM at 48, and #OCC/# is selected as the system for restarting at 49.
0CIICI and 60MM3 are connected and the system is restarted.

〔Effect of the invention〕

As described above, according to the present invention, it is possible to detect a faulty device before restarting, and the system can be restarted quickly and accurately, which contributes to improving system reliability. big.

3. . . #0 memory (MM), 4. . . #
l memory (MM).

[Brief explanation of the drawing]

FIG. 1(a), (bl is a functional block diagram of the present invention,
2 is a flowchart of the first embodiment of the system configuration determination method; FIG. 3 is a flowchart of the second embodiment of the system configuration determination method; FIG. 4 is a diagram showing an embodiment of system restart processing execution; FIG. 5 is a diagram showing a rotation method as a conventional example of a system configuration determination method. FIG. 6 is a diagram showing a conventional example of a flowchart of a system configuration determination method. FIG. 7 is an explanation of problems with the rotation method. It is a diagram. 1...#0 central controller (CC)/channel controller (CHC), 2...#1 central controller (CC)/channel controller (CHC),

Claims (1)

[Claims] 1) In the restart operation of an electronic exchange system in which a central control unit, a channel control unit, and a storage device are duplicated, the configuration of the active system before restart among the duplicated devices (20), starts a system restart operation (21), and determines a system configuration at the time of restart without incorporating the configuration of the current system before the restart (22). Configuration determination method. 2) In the restart operation of an electronic exchange system in which the central control unit, channel control unit, and storage device are duplicated, the restart operation of the system is started (21), and the above-mentioned 2. Determine which device caused the failure among the multiplexed devices (2)
3) A system configuration determination method at the time of restart operation, characterized in that the system configuration at the time of restart is determined (24) without incorporating the device that caused the failure.