JPS62198943A - Fault information collecting and analyzing system for electronic computer system - Google Patents

Abstract

PURPOSE:To improve to overall reliability and the continuation of actions for the titles system by starting the analysis of the collected and unanalyzed fault information according to the prescribed priority based on the fault codes or the before-afater relation of information collection. CONSTITUTION:A fault information collection/analysis control part 11 receives a request for collection of fault information received from one of processors 21-2n and shift the control to a fault information collecting part 12 to start collection of the fault information. Thus the part 12 collects the fault information via interfaces 41-4n and adds the collection identification codes to these collected information to store them in a fault information store part 14 and also to store a list of collection identification codes or fault codes in a code store part 15. Then the part 11 reads out the contents of the part 15 and decides the fault information to be analyzed next based on the prescribed priority. The the part 11 delivers the control to a fault information analyzing part 13 while designating the the fault information to be analyzed and reads otu the designated fault information to inform it to an operator after analysis.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a fault information collection and analysis method used in electronic computer systems.

A conventional electronic computer system is composed of a combination of a large number of processing devices such as a central processing unit, an input/output control device, and peripheral control devices. In addition, in computer systems, each processing unit is duplicated so that the operation of the entire system does not stop even if a failure occurs in one processing unit, and a server that performs failure diagnosis in parallel with the operation of the entire system. A system diagnostic device called a processor is now being added.

Conventionally, in the system diagnostic equipment described above, when a failure occurs in a certain processing device, failure information is collected from that processing device, and as soon as this collection is completed, the system immediately moves to analyze the failure information to identify the location of the failure. It is output as a result.

Problems to be Solved by the Invention 1: In conventional system diagnostic devices, the collection and analysis of fault information are performed as a series of continuous operations, so when a fault occurs in a certain processing device, the collection of fault information is stopped. Even if a failure occurs in another processing unit immediately after starting, the collection and analysis of failure information for the other processing unit will be unconditionally delayed until the analysis of the immediately preceding failure processing unit is completed. It turns out.

Therefore, even if the subsequent failure is a serious failure that requires more urgent diagnosis than the earlier failure, the system will have to wait until diagnosis of the earlier minor failure has been completed, which will reduce the reliability of the entire system. There are problems in terms of performance and mVt characteristics of operation.

Means for Solving the Constituent Problems of the Invention The fault information collection/analysis method of the present invention that solves the problems of the prior art described above prioritizes collection of fault information from each processing device before starting its analysis. , saves each collected failure information while displaying the collection relationship, and if there is multiple unanalyzed failure information being saved, the failure code and the collection relationship regarding the type and scope of the failure included in each. The system is configured to increase the reliability and continuity of operation of the entire system by starting analysis according to a predetermined priority determined from the following.

Hereinafter, the operation of the present invention will be explained in detail together with examples.

Embodiment FIG. 1 is a block diagram showing the configuration of a computer system to which a fault information collection/analysis method according to an embodiment of the present invention is applied.

This electronic computer system has a central processing unit,
Collects and analyzes fault information from multiple processing devices 21, 22...2n consisting of input/output control devices, peripheral control devices, etc. and the one in which a fault has occurred among these processing devices, and identifies fault locations. System diagnostic device 1 that outputs results
Consists of 0.

The system diagnostic device 10 includes a fault information collection/analysis control section 1
1, a failure information collection unit 12, a failure information analysis unit 13,
It includes a fault information storage section 14 and a code storage section 15.

The failure information collection/analysis control unit 11 connects signal lines 31 to 3n from any of the processing devices 21 to 2n of the system in operation.
When receiving a request to collect failure information due to the occurrence of a failure via the failure information collection unit 12, control is passed to the failure information collection unit 12 to start collecting failure information. The fault information collection unit 12 to which control has been passed collects fault information from the corresponding fault occurrence processing device via the diagnostic interfaces 41 to 4n, adds a collection identification code indicating the collection relationship, and identifies the fault. Information storage section 14
At the same time, a list of the collected identification codes and fault codes included in the collected fault information is stored in the code storage unit 15.

The failure information collection/analysis control unit 11 reads the contents of the code storage unit 15 and determines the failure information to be analyzed next in accordance with a predetermined priority order, at a time when there is no request to collect failure information. , and passes control to the failure information analysis unit 13 while specifying this. Failure analysis unit 13 to which control has been handed over
reads out the designated failure information from the failure information storage unit 14, analyzes it, and notifies the system operator of the analysis result, which includes pointing out the failure location.

FIG. 2, FIG. 3, and FIG. 4 respectively show an example of the control procedure by the fault information collection/analysis control unit 1, the fault information collection unit 12, and the fault information analysis unit 13 in the system diagnosis bag W10. It is a flowchart.

When the failure information collection/analysis control unit 11 starts processing,
In the first step lla of FIG.
By scanning 1 to 3n, the presence or absence of a failure information collection request is checked. If there is a fault information collection request, the fault information collection/analysis control unit 11 proceeds to step llb, starts the fault information collection unit 12, notifies the collection target, and then
Proceeding to step 11c, the process transitions to a waiting state for a collection completion notification from the failure information collection unit 12.

The fault information collection unit 12 started by the fault information collection/analysis control unit 11 proceeds to the first step 12a in FIG.
The specified failure information is sent to the corresponding location f'! ! Device 21~
Collect from one of 2n. This collection of fault information includes reading fault codes that display the type and scope of the fault.
It consists of collecting various data at the time of failure based on the known shift-out method. In other words, a serial transfer path is formed so as to connect predetermined bit positions of a predetermined register that is built into the relevant processing device and holds parallel data during normal operation, and a shift register in the diagnostic device is connected in synchronization with the system clock. A shift-out operation is performed in which the data is read within the memory.

When the failure information collection unit 12 finishes collecting the specified failure information, it proceeds to the next step 12b, and adds a collection identification code such as a serial number and an identification code of the failure processing device to the collected failure information to create failure information. It is stored in the storage unit 14. The fault information to which these identification codes are attached, as illustrated in Figure 5, consists of a fault code that displays the type of fault and the range of the fault, and detailed information collected by the shift-out operation described above. There is.

When the fault information collection unit 12 finishes storing the collected fault information in the fault information storage unit 14, it proceeds to the next step 12c and stores the collection identification code of the collected fault information, the identification code of the processing device, and the fault code in the code storage unit 15. Store. In the final step 12d, the failure information collection unit 12 notifies the failure information collection/analysis control unit 11 of the end of failure information collection, and ends the process.

Failure information collection/receiver notification of completion of failure information collection
The analysis control unit 11 returns from step 11c in FIG. 2 to step lla in the same figure, checks whether there is a request to collect new failure information, and if it exists, returns to step 11b.
and llc, repeat the above operations.

If there is no request to collect new fault information, the processing of the fault information collection/analysis control unit 11 proceeds to step l.
Move from la to step lid. In step lid, the failure information collection/analysis control unit 11 checks whether there is a collection identification code stored in the code storage unit 15, and if it does not exist, the process returns to step lla. When the fault information collection/analysis control unit 11 detects the existence of the collection identification code being stored, it proceeds to step lie, reads all the codes from the code storage unit 15, and starts analyzing the fault information first according to a predetermined priority order. Select one piece of information and start the fault information analysis section 13.

The priority order for starting this analysis is determined based on the content of the failure code and the collection identification code that indicates the relationship between collection. In other words, the type of failure (intermittent failure, fixed failure, bit error, etc.) and range (
(the entire processing device, a part of the processing device, etc.), one piece of fault information about the most serious fault is selected. If there are multiple failures with the same severity, the first piece of failure information collected is selected based on the collection identification code. When there is only one piece of fault information, that fault information is unconditionally targeted for analysis.

The failure information collection/analysis control unit 11 includes a failure information analysis unit 13
When the system starts up and notifies the collection identification code of the failure information to be analyzed, the process advances to step 11f and shifts to a waiting state for an analysis completion notification from the failure information analysis unit 13.

The activated failure information analysis unit 13 performs the first step 13.
In step a, the failure information corresponding to the designated collection identification code is read from the failure information storage unit 14, analyzed, and the operator is notified of the analysis result, which includes pointing out the failure location. When this analysis is completed, the failure information analysis unit 13 instructs the failure information collection/analysis control unit 1 to terminate the analysis in step 13b.
After notifying 1, all processing ends.

The fault information collection/analysis control unit II is the fault gR@analysis unit 1.
When the notification of the end of the analysis is received from step 3, the process proceeds to step 11g, where it is determined whether or not there is a command to end the process, and if there is no command, the process returns to step 11a and the above-described control is repeated.

Above, we have explained the configuration in which after the analysis of failure information starts, requests for collection of subsequent failure information are not accepted until the analysis is finished. The configuration is such that the analysis of the information is interrupted and the collection is completed, and if the subsequent failure is more serious than the failure being analyzed, the analysis of the earlier failure information is postponed and the analysis of the subsequent failure information is started. You can also.

Effects of the Invention As explained in detail above, the failure information collection and analysis method according to the present invention prioritizes the collection of failure information from each processing device before starting its analysis, and collects unanalyzed collected information with priority. Since the configuration starts analysis of failure information according to a predetermined priority order based on the failure code and the relationship between collection priority, analysis of failure information is completed in order of urgency.
The effect is that the reliability of the entire system and the wlVt performance of the operation are greatly improved.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of a computer system to which a fault information collection/analysis method according to an embodiment of the present invention is applied, and FIGS. 2 to 4 respectively show the diagnosis device 10 of FIG. A flowchart illustrating the operations of the fault information collection/analysis control section 11, the fault information collection section 12, and the fault information analysis section 13 in FIG. FIG. 3 is a format diagram showing the structure of failure information. DESCRIPTION OF SYMBOLS 10: System diagnostic device, 11: Failure information collection/analysis control unit, 12: Failure information collection unit, 13: Failure information analysis unit, 14: Failure information storage unit, 15: Code storage unit.

Claims (1)

[Claims] In a fault information collection/analysis method for a computer system that collects and analyzes fault information from a plurality of processing devices constituting the computer system, collection of fault information from each processing device is performed in the analysis. Priority is given to the start of the process, and each collected failure information is saved while displaying the relationship between collections, and if there is multiple unanalyzed failure information being saved, failure information related to the type and range of failure included in each is saved. A fault information collection/analysis method for a computer system, characterized in that analysis is started according to a predetermined priority determined from a code and a collection sequence.