JP2560892B2 - Error information processing device - Google Patents

Description

[Description] Regarding the processing of error information in a computer, an object of the error information processing device is to generate error information so that the information can be easily associated with each other. And an error processing unit, wherein the error management unit receives an identifier generation request from the request source at the time of error to generate an identifier for identifying each error, and generates an identifier having a different value for each request, Returning to the request source, the error processing unit is provided for each of a required program and a program called by the required program, and each program has an identifier holding unit, and the error holding unit is held in the identifier holding unit. As an identifier, when the error detection is an initial error that has no secondary relationship with other programs, a value obtained by requesting the error management unit is held and the initial error is held. In the case of a secondary error accompanying the error, the value notified as the identifier from the called program is retained, the error is detected during the predetermined processing of the program, and the error processing related to the detection of the error is performed. When executing, the identifier stored in the identifier storage unit is added to the diagnostic information output by the error processing and output, and when the calling source program exists, the identifier is stored in the calling source. Configure to notify the program.

[Industrial applications]

The present invention relates to error information processing in a computer, and more particularly to an error information processing device that generates error information so that information can be easily associated with each other.

[Conventional technology]

FIG. 5 is a block diagram showing a configuration example of a computer system. When any error is detected during execution of the program B when the program 1 (referred to as the program A) of the program executed in the computer system calls and executes the program 2 (referred to as the program B), After the error processing unit 8b provided in the program B performs necessary error processing and notifies the detected error event to the error management unit 3 that centrally manages the error event of the system as an event notification b (in the figure, The control is returned to the program A along with the path (hereinafter, simply referred to as “the same case”) and the return information indicating the failure ().

Therefore, the program A performs a process corresponding to the occurrence of an error, but when an error is detected as a secondary result, the error processing unit 8a of the program A causes an error event (event notification a) in the error management unit 3. To notify ().

In the error management unit 3, the log output unit 6 stores the received error event notification record in the log file 5 and outputs it to the terminal 4 as an error message. The contents of the log file 5 are separately analyzed by a failure analysis program or the like.

In that case, if the event notification b of the original error notified by the program B and the event notification a of the secondary error notified by the program A as described above are present, on the record of the log file 5, If the relation between those event notification records is specified, the processing becomes easy.

Therefore, Japanese Patent Application No. Sho 62-88143 “Unified error event centralized management system” filed by the present applicant discloses the following system. That is, the error management unit 3 is added to the identifier generation unit 7
Each time an event notification is received, a new number value that has been progressively generated is generated as an identifier and returned to the event notification source, and it is also passed to the log output unit 6 to be added to the event notification record that has just been received. To do.

In the programs A and B, for example, when the program B detects an error and notifies the event notification b as described above, when the identifier N1 is received, and when the program returns to the calling program A, The identifier N1 is also added to and passed. Therefore, when the program A notifies the event notification a as if an error was detected, the event notification a
The identifier N1 is set in the field of the related event provided in the record and passed.

On the other hand, the identifier generation unit 7 generates and returns the next new identifier N2 (), and also passes this identifier N2 to the log output unit 6 so as to add it to the event notification.

As described above, in the event notification record output by the log output unit 6, as shown in FIG. 2A, in addition to various information of the conventional event notification as illustrated in the figure, a field of the event identifier A column of related event identifiers is added, and a chain of identifiers recorded in both columns as shown in the figure makes it relatively easy to associate a series of events.

[Problems to be Solved by the Invention]

According to the above-mentioned technique, a series of related events can be recognized by an identifier, but it takes a relatively long time to search a series of events because it is a chain, and when identifying individual events. However, since it cannot be recognized that it belongs to a series of event groups unless the chain is traced, there is a problem that it is inconvenient for individual processing.

An object of the present invention is to provide an error information processing device that generates error information so that information can be easily associated with each other.

[Means for solving the problem]

 FIG. 1 is a block diagram showing the configuration of the present invention.

The figure shows the configuration of the error information processing apparatus, which has an error management unit 12 and error processing units 18a and 18b.
12 is an identifier generation unit 17, a log output unit 16, and a log reference unit
19, and the identifier generation unit 17 receives the request for the identifier,
An identification sheet having a different value is generated for each request and returned to the request source. The log output unit 16 stores the predetermined information output from the error processing units 18a and 18b in the log file 15 and refers to the log. The part 19 is for the log file 15 when it receives a predetermined designation.
The information stored in the key error processing unit 18a, 18b for the value of the identifier is provided for each required program and each program called by the required program, and has an identifier holding unit in each program, As the identifier held in the identifier holding unit, when the error detection is an initial error having no secondary relationship with other programs, a value obtained by requesting the error management unit 12 is held, In the case of a secondary error accompanying the initial error, the value notified as the identifier from the called program is held, and the error is detected during the predetermined processing of the program, and is related to the detection of the error. When executing error processing, the identifier held in the identifier holding unit is added to the diagnostic information output by the error processing and output, and the calling program Is present, the identifier is notified to the calling program.

The error processing units 18a and 18b add the identifier and the root cause display to the diagnostic information to be output, and when the root cause display is acquired by requesting the identifier to the error management unit 12. The value to be output is different from the value to be output when the identifier is notified by another program.

[Work]

According to the error information processing apparatus of the present invention, a group of related event notifications has the same key identifier, and error information other than the event notification, such as diagnostic information, can have the same identifier. Furthermore, since it is possible to distinguish the event notification of the first event caused by the root cause display from other event notifications, the related event notifications can be easily searched in the log reference section, and By adding search conditions, it is possible to efficiently perform the work of taking out a notification of a specific event in the group and collating it with the diagnostic information.

〔Example〕

In FIG. 1, when the program 10 (program A) calls and executes the program 11 (program B),
When some error is detected during execution of the program B, the error processing unit 18b provided in the program B performs necessary error processing and notifies the detected error event to the error management unit 12 as an event notification b. (), And returns control to program A with return information indicating a failure ().
However, if the program B is, for example, the lowest control program, an identifier request is issued to the identifier generation unit 17 of the error management unit 12 immediately after an error is detected, and for example, the identifier N1 is acquired ().

After acquiring the identifier in this way, the error processing unit 18b collects, for example, predetermined diagnostic information, and when writing it to the diagnostic information file 13, adds the identifier N1 as key information and outputs (), The record of the event notification b is also provided with a field of the event association identifier, the identifier N1 is set, and the field of the root cause display is set to, for example, "1" to display the event of the root cause. Output.

The program A to which the control is returned performs processing corresponding to the error occurrence, but when the secondary error as a result is recognized, the error processing unit 18a of the program A causes the error management unit 1 to
Notify 2 of the error event as event notification a (). At that time, instead of requesting a new identifier as in the past,
Identifier N1 passed as return information from program B
Is set in the field of the event association identifier of the event notification a as in the case of the event notification b, but the event of the subordinate cause is displayed by setting the field of the root cause display to, for example, “0”. Output.

In the error management unit 12, the log output unit 16 stores the received error event notification record in the log file 15 and outputs a message to the terminal 4.

As described above, in the event notification record output by the log output unit 16, as shown in FIG. 2B, in addition to the fields of various information of event notification, the field of event association identifier and the field of root cause display are included. Since a column of event association identifier is also provided in the record of diagnostic information, as shown in the figure, it is relatively easy to associate a series of events and related diagnostic information by matching the identifiers in the column of event association identifier. As a result, the log reference unit 19 can quickly search using the event association identifier field as a key.

FIG. 3 is a block diagram showing one embodiment of the present invention.
When the user program 20 calls the file management program 21 to access a file on the magnetic disk storage device 29, the file management program 21 is configured to receive the service of the input / output supervisor 22 for controlling the magnetic disk storage device 29. ing. The hardware configuration management program 23 is a mechanism that manages the system configuration in response to a failure of the hardware device, and substantially handles errors as a part of the error processing unit of the input / output supervisor 22 as described later. To do.

When the I / O supervisor 22 detects a failure of the magnetic disk storage device 29 during execution, the error processing unit 18c of the I / O supervisor requests the error management unit 12 for an identifier (), and obtains, for example, the identifier N1. Then, the hardware configuration management program 23 is attached with the identifier N1 and a hardware failure notification is sent (), and then the file management program 21 of the calling source is returned to. Identifier N1 as return information in that case
give.

The hardware configuration management program 23 attaches diagnostic information to the magnetic disk storage device 29 based on the failure notification from the I / O supervisor 22 and assigns the identifier N1 (FIG. 2 (c)).
Diagnostic information file as a record as shown in
Since the event notification having the contents shown in 13 is sent to the error management unit 12 (), the log output unit 16 causes the log file 15
Store to.

In the file management program 21, I / O supervisor
Upon knowing that the processing of the file could not be completed due to a hardware failure of the magnetic disk storage device 29 due to the abnormal recovery from 22, the error processing unit 18d notifies the error management unit 12 of an error event notification regarding the file ( ), Is stored in the log file 15. The event notification at this time has the content shown in FIG.

FIG. 4 is a diagram showing an example of the processing flow of the system shown in FIG. 3, and the file management program 21 buffers the write record in processing step 30 in response to a request from the user program 20, and Step 31
If the file needs to be updated, the input / output supervisor 22 is called in process step 32 to issue a file write request.

After that, when returning from the input / output supervisor 22, the return information is judged in the processing step 33 whether the processing is completed. For example, if the returning value is 0, the processing is completed and the operation is normally returned to the calling source. Then, the event notification is sent to the error management unit 12 and then the error is recovered. In this case, the event notification has, for example, the content shown in FIG. 2C, and the value of the return value (non-zero as described above) passed from the input / output supervisor 22 is used as the identifier in the event association identifier column. The cause display field is 0.

The I / O supervisor 22 is a file management program 21.
Then, in step 40, a required I / O command is issued to operate the magnetic disk storage device 29. In processing step 41, the execution result of the I / O instruction is judged, and if it is completed normally, the return value passed as the return information is set to 0 and the operation is returned to normal.

If the instruction cannot be completed due to some hardware failure, the error management unit 12 is requested in process step 42 to acquire the identifier (value is N1), and the failure notification is sent to the hardware in process step 43. Notify the configuration management program 23. This failure notification is made up of an identifier N1, a device identification name, a failure type, and so on. After that, in the processing step 44, the identifier N1 just acquired is set as the return value, and abnormal return is made to the calling source.

When the hardware configuration management program 23 receives the failure notification from the input / output supervisor 22, it makes a diagnostic information record as shown in FIG. The diagnostic information data is composed of a device identification name, a fault type, sense information, and the like.

Then, in process step 61, the failed magnetic disk storage device 29
Is separated from the system configuration so that it will not be used, and an event notification as shown in FIG. The event-specific data of the event notification includes a device identification name, a record position at the time of failure occurrence, and other failure statuses.

In the error management unit 12, when an identifier is requested, the identifier generation unit 17 generates a new value in processing step 50 and returns it as an identifier to the request source. For the value of the identifier, use a method such as incrementing the number by +1 for each request or synthesizing from the requester's identification name and the request date and time (for example, second display), and make it unique to the system within the required period. To be

Upon receiving the event notification, the log output unit 16 performs the processing step 51.
Assemble a record according to its contents and process step 52
And store it in log file 15. In processing step 53, the necessity of message output (for example, designated in event notification) is identified, and if necessary, an error message is generated from the event notification and output in processing step 54.

〔The invention's effect〕

As is clear from the above description, according to the present invention, in the processing of error information of a computer, it is possible to generate error information so that the information can be easily associated with each other, so that the work efficiency of error analysis processing and the like is improved. It has a remarkable industrial effect.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of the present invention, FIG. 2 is a diagram explaining an example of error information, FIG. 3 is a block diagram showing an embodiment of the present invention, and FIG. 4 is a flow chart of processing of the present invention. FIG. 5 is a block diagram showing a conventional configuration example. In the figure, 1, 2, 10, 11 are programs, 3, 12 are error management sections, 5 and 15 are log files, 6 and 16 are log output sections, 7 and 17 are identifier generation sections, and 8a, 8b and 18a ... 18d is an error processing unit, 13 is a diagnostic information file, 19 is a log reference unit, 20 is a user program, 21 is a file management program, 22 is an input / output supervisor, 23 is a hardware configuration management program, 29 is a magnetic disk storage device, 30 to 34, 40 to 44, 50 to 54, 60 to 62 indicate processing steps.

Claims (3)

(57) [Claims] 1. An error management unit and an error processing unit, wherein the error management unit receives an identifier generation request for identifying an individual error from a request source at the time of an error and receives the request for each request. An identifier having a different value is generated and returned to the request source, the error processing unit is provided for each required program and each program called by the required program, and each program has an identifier holding unit, As the identifier held in the identifier holding unit, when the error detection is an initial error having no secondary relationship with another program, a value obtained by requesting the error management unit is held, In the case of a secondary error accompanying the initial error, the value notified as the identifier from the called program is retained, and the error is detected during the predetermined processing of the program, and the error When executing the error processing related to the detection, the diagnostic information output by the error processing of the identifier held in the identifier holding unit is added and output, and when the calling program exists, An error information processing apparatus, characterized in that the identifier is notified to a calling program. 2. The error processing unit adds the identifier and a root cause display to the diagnostic information to be output, and outputs the root cause display when the identifier is requested from the error management unit and acquired. The error information processing apparatus according to claim 1, wherein the value to be output is different from the value output when the identifier is notified from another program. 3. The error management unit includes a log output unit and a log reference unit, the log output unit stores diagnostic information output from the error processing unit in a log file, and refers to the log. Part, the information stored in the log file,
The error information processing apparatus according to claim 1, wherein the error information processing apparatus is configured to search and output using the value of the identifier as a key.