JPH07262054A - Failure information management system - Google Patents

Abstract

PURPOSE:To provide a failure information management system which can perform various types of management of threshold value in accordance with the types and importance degree of errors and also corresponding to the system operations. CONSTITUTION:When a failure occurs, an error code is produced and stored in a detailed error log file. At the same time, the error code and the error occurrence frequency (e) are inputted to a threshold value management part 5. The part 5 reads a threshold value management table 10a out of a threshold value management file. The table 10 a contains the entries for management cycles (c), lower limit value (l), upper limit value (u), error mask code, error comparison codes, etc. Then, a logical arithmetic operation is performed between the error code and the error mask code and also the error code is compared with the error comparison code, and the type of the corresponding error is specified and then judged by means of a conditional expression, [1<=(e mode c)<=u]. When this expression is satisfied, it is decided that the failure shown in the error code is out of its allowance range. In such a constitution of a failure information management system, the decided error is reported to a maintenance center.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a fault information management technique used for improving the reliability and maintainability of an information processing device.

[0002]

2. Description of the Related Art For example, in an information processing apparatus, an error code created when a failure occurs is indispensable for system maintenance and operation. This error code is diversified as the system becomes large-scale and complicated. In addition, the failure that occurs affects a wide range, and it takes a lot of time to investigate the cause. Therefore, it is necessary to deal with these in the creation and storage of error codes and preventive maintenance.

As a conventional fault information management system by threshold management, the technique described in Japanese Patent Laid-Open No. 3-250227 has been known. That is, it has a storage table having a plurality of storage records composed of a failure code and a failure counter, and a reference table having a plurality of reference records composed of a failure code and a threshold, and sequentially references the reference records of the reference table to store the storage records. If the fault codes of
The failure counter is compared with the threshold, and if the failure counter is larger than the threshold, a failure code is transmitted to the monitoring host and the failure counter is reset. On the other hand, when there is no reference record having the same fault code in the accumulated record, the fault code is transmitted to the monitoring host and the fault counter in the accumulated record is reset.

[0004]

According to the above-mentioned prior art, in the fault information management system by the fault code threshold management, all possible values of the fault code which is the threshold management target are registered in the reference table. As it is necessary, the reference table will increase. Further, since there is only one threshold value registered in the reference record of the reference table corresponding to the fault code that is the target of threshold value management, it is determined that the threshold value is exceeded every time a fault occurs as shown in FIG. Manage or Figure 4
If the failure count is a multiple of the set threshold value when a failure occurs as in (f), the method is limited to the threshold value management method of managing that the threshold value is exceeded. Therefore, for example, if you want to check the frequency of garbled data on the line, you can check the frequency by exceeding the threshold once at a certain number of times and sending it to the monitoring host, but you cannot perform such threshold management. .
Therefore, in order to manage the threshold value such that the threshold value is exceeded only once at a certain number of times, and then the threshold value is not exceeded,
Using the threshold management of (f), after the threshold has been transmitted to the monitoring host, the reference record corresponding to the fault code that is the threshold management must be deleted from the reference table. Information is lost and operation becomes complicated. If you do not delete it, after resetting the failure count,
Once again, the inconvenience of exceeding the failure count occurs again.

An object of the present invention is to provide a fault information management system capable of performing various threshold managements according to the type and importance of errors and system operation for preventive maintenance.

Another object of the present invention is to provide a fault information management system capable of simplifying error code management.

[0007]

FIG. 2 is a block diagram showing an example of the principle of threshold value management in the fault information management system of the present invention.

When an error occurs in the information processing device, an error code is created, and the error code indicates the error occurrence count (e).
The threshold management unit receives the information added with. Threshold management unit 5
Threshold management table 10a used by the user to perform threshold management
Is a configuration having a plurality of reference unit records, a management cycle (c) indicating an interval of error occurrence times of a failure for which threshold management is performed, a lower limit value (l) of error occurrence times of a failure for performing threshold management, and a threshold value. The upper limit value (u) of the error occurrence count of the fault to be managed, the error mask code that is the mask value of the error code, and the error comparison code that compares the logical product of the error code and the error mask code or the value of the logical sum operation result And The threshold management unit sequentially refers to the reference unit record of the threshold management table, calculates the logical product or logical sum of the error code and the error mask code, and when the operation result and the error comparison code match, the error code is the threshold management target. It is determined that the number of error occurrences is modulo the management cycle of the reference unit record, and if this value is greater than or equal to the lower limit value and less than or equal to the upper limit value, [l ≦ (e mod
c) ≦ u], it is determined that the failure exceeds the allowable range. On the other hand, if there is no match between the error comparison code and the result of the logical product or the logical sum of the error mask codes of all the reference unit records in the threshold management table, it is determined that the error code is not the threshold management target. .
In addition, it was determined that the failure was within the permissible range if it was determined as a threshold management target, but the value of the number of error occurrences modulo in the management cycle of the reference unit record is not less than the lower limit value and not more than the upper limit value. To do.

[0009]

According to the present invention, various threshold management methods as shown in FIGS. 4 (a) to 4 (g) can be performed by combining the lower limit value, the upper limit value, and the value of the management cycle, and the type of error and Threshold management of various failures can be performed according to the degree of importance and system operation. Further, the content of the error code is hierarchized according to the hardware and the classification of the failure, and the failure is separated by the error mask code, whereby the increase of the threshold management table can be prevented and the capacity of the threshold management table can be reduced.

[0010]

Embodiments of the present invention will now be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing an example of the configuration of an information processing system in which a failure information management system according to an embodiment of the present invention is implemented.

In FIG. 1, reference numeral 1 is a detailed error log collection unit for collecting different detailed error logs depending on the failure when the failure occurs and storing it in the detailed error log file 8. Reference numeral 2 is a detailed error log showing various failures in a unified format. An error code creation unit for creating the expressed error code, 3 is an error code creation table used by the error code creation unit 2 for creating the error code, 4 is an error code storage unit for storing the error code in the error code file 9, and 5 is Threshold management is performed by the data obtained by adding the number of error occurrences to the error code and the threshold management table 10a stored in the threshold management file 10 having the configuration shown in FIG. A threshold management unit for storing threshold excess information, 6 is a detailed error log file 8
Output of stored data of the above, deletion of stored data, output of stored data of the error code file 9, deletion of error code, input / output and deletion of data of the threshold management table 10a of the threshold management file 10, threshold excess information file 11 Error information input / output unit for outputting threshold excess information data or deleting threshold excess information data via an input / output device 7 such as a keyboard or a CRT; 12 is an error code storage unit 4; When it is detected that the specified value is exceeded, or when the threshold value management unit 5 detects that the error code threshold value is exceeded, the communication device 13 is controlled to control the maintenance center 14.
The communication control unit 12 sends each information to the.

FIG. 3 is a flow chart showing an example of error code logging processing in the fault information management system of this embodiment. The error code logging process from the occurrence of a failure will be described below with reference to FIGS. 1, 2 and 3.

(1). When a failure occurs, the detailed error log collection unit 1 collects a detailed error log having different collection contents and log data length depending on each failure and stores it in the detailed error log file 8. Further, the detailed error log collected is passed to the error code creation unit 2 (step 100).

(2). The error code creation unit 2 creates an error code that represents each failure in a uniform format using the detailed error log received from the detailed error log collection unit 1 and the error code creation table 3, and stores it in the error code storage unit 4. Hand over (step 101).

(3). The error code storage unit 4 stores the error code created by the error code creation unit 2 in the error code file 9, and passes the data obtained by adding the error occurrence count to the error code to the threshold value management unit 5 (step 10).
2).

(4). The threshold management unit 5 searches for a reference unit record in which the valid / invalid flag of the reference unit record of the threshold management table 10a of the threshold management file 10 is valid (step 200).

(5). It is checked whether or not all the valid reference unit records have been searched, and if all the valid reference unit records have been searched, the processing ends (step 201).

(6). When the valid reference unit record is found, the logical product operation of the error code received from the error code storage unit 4 and the error mask code of the valid reference unit record is performed (step 202).

(7). When the value of the logical product operation result of the error code and the error mask code of the valid reference unit record and the error comparison code of the valid reference unit record do not match, the process returns to (4) (step 200) (step 203). .

(8). When the logical product operation result of the error code and the error mask code of the valid reference unit record matches the error comparison code, the modulo of the management cycle of the valid reference unit record with respect to the number of error occurrences is calculated (step 204).

(9). If the value obtained by calculating the modulo of the management cycle of the valid reference unit record with respect to the number of error occurrences is not more than the lower limit value and not more than the upper limit value of the valid reference unit record, the number of error occurrences of the error code does not exceed the threshold value (failure is It is determined to be within the allowable range), and the processing ends (step 205).

(10). If the value obtained by calculating the modulo of the management cycle of the effective reference unit record with respect to the number of error occurrences is greater than or equal to the lower limit value and less than or equal to the upper limit value of the effective reference unit record, it is determined that the number of error occurrences of the error code exceeds the threshold value. The threshold excess information is stored in the threshold excess information file 11 (step 206).

(11). The communication control unit 12 checks whether or not the information processing apparatus is connected to the maintenance center 14 by a communication line or the like, and if it is not connected to the maintenance center 14,
The process ends (step 207).

(12). The information processing device is a maintenance center 1
4 is connected by a communication line or the like, the threshold value excess information stored in the threshold value excess information file 11 is controlled by the communication device 13 and transmitted to the maintenance center 14 to end the processing (step 208).

FIG. 4 is a diagram showing an example of various threshold value management methods by the combination of the lower limit value (l), the upper limit value (u) and the management cycle (c) which are the parameters of the threshold value management table 10a in this embodiment. is there. Hereinafter, a threshold management method and an example of a failure for which it is effective will be described with reference to FIG.

In FIG. 4A, l = 0, u = i-1, and c = i.
This shows the threshold value management method in the case of (i> 0), and management is performed to determine that the allowable range is exceeded each time a corresponding failure occurs. An example of a failure that uses this threshold value management method is a serious failure that requires immediate countermeasures.

In FIG. 4B, l = 0, u = i, c = ∞ (i
> 0) represents the threshold value management method, and performs management to determine that the allowable range has been exceeded up to i times from the occurrence of the corresponding failure.

In FIG. 4C, l = i, u = ∞, c = ∞ (i
> 0) represents a threshold value management method, and performs management to determine that the number of occurrences of the corresponding failure exceeds the allowable range after i times. An example of a failure that uses this threshold value management method is a failure that was not a serious failure during system operation, but became a severe failure due to a system change or the like after operation.

In FIG. 4D, l = i, u = i, and c = ∞ (i
> 0) represents a threshold value management method, in which management is performed to determine that the number of occurrences of the corresponding failure exceeds the permissible range only when i times. As an example of a failure using this threshold value management method, there is a case where a frequency investigation of garbled data failure of a line is performed.

In FIG. 4E, l = i, u = j, c = ∞ (0
It represents a threshold value management method in the case of <i <j), and performs management to determine that the number of occurrences of the corresponding failure is beyond the allowable range from i times to j times.

FIG. 4 (f) shows a threshold value management method in the case of the lower limit value = 0, the upper limit value = 0, and the management cycle = n (n> 0). The number of times of occurrence of the corresponding failure is a multiple of n. In this case, control is performed to determine that the allowable range is exceeded. An example of a failure using this threshold value management method is a failure in which the failure occurrence frequency and failure countermeasures such as component replacement are already known, and it is sufficient to transmit at a multiple of a certain number of times.

FIG. 4G shows a threshold value management method in the case of the lower limit value = i, the upper limit value = j, and the management cycle = n (0 <i <j <n). When the number of times is n or more and j times or less in a cycle of n times, management is performed to determine that the allowable range is exceeded.

FIG. 5 is a conceptual diagram showing an example of error code system and threshold management grouping in this embodiment.

FIG. 5A is a diagram showing an error code for each unit of the apparatus. The unit is divided into a CPU, a bus adapter, a device, etc., and is identified by the first 4 bits of the error code.

FIG. 5B is a diagram showing an error code for each device. The device is a cassette magnetic tape unit (C
MT), digital audio tape device (DAT), hard disk device (HDD), etc., and is identified by the first byte of the error code.

FIG. 5C is a diagram showing error codes of the major categories of DAT failures. It is divided into timeout failure, adapter failure, device failure, retry failure failure, etc., and is identified by the first 2 bytes of the error code.

FIG. 5D is a diagram showing a small classification of DAT retry failure failures. The retry failure failure is divided into "tape loading is not completed", "reception command parameter has invalid field", etc., and is identified by the first 6 bytes of the error code.

In the case of the error code system shown in FIGS. 5A to 5D, for example, when a failure occurs such that the tape load is not completed due to a DAT retry failure,
Device, DAT, DAT retry failure failure,
It is possible to perform hierarchical grouping of failures such that tape loading is not completed (no grouping) due to DAT retry failure, and one reference unit record of the threshold management table 10a shows the error mask code and error shown in FIG. 5 (e). This can be done by setting the comparison code.

As described above, according to this embodiment, the failure analysis is performed by creating an error code representing each failure in a unified format from the detailed error log in which the collection contents and the log data length are different depending on each failure. The procedure is simplified and the time required for troubleshooting can be shortened. Also,
Even if the failure analysis cannot be performed only with the error code, the detailed error log is stored separately from the error code, so that more detailed failure analysis can be performed. Also, by using the upper limit value, the lower limit value, the management cycle, and the number of error occurrences as the threshold value for performing the error code threshold value management, the threshold value management method can be changed, and the type of failure, the importance level,
And various management according to system operation is possible.

Further, the error codes to be threshold-controlled are grouped by fault masking using the error mask code and the error comparison code, and the threshold management table 1
The reference unit record of 0a is provided with a valid / invalid flag,
Since the reference unit record can be deleted, the capacity of the threshold management table 10a can be reduced. Further, detailed error log data which is failure information, an error code, a threshold value management table 10a, and threshold value excess information can be input / output by the input / output device 7 and the failure information input / output unit 6, so that the system operator can grasp the failure information in a timely manner. And so on.

[0042]

As described above, according to the fault information management system of the present invention, the upper limit value, the lower limit value, the management cycle, and the number of error occurrences are used as the threshold values for the error code threshold value management. The threshold management method can be changed in various ways, and various managements can be performed according to the type and importance of failure and system operation. Further, by grouping the error codes to be threshold-managed with the error mask code and the error comparison code, the failures are hierarchized to prevent an increase in the threshold management table and reduce the capacity of the threshold management table. In addition, the information processing device is connected to the maintenance center by a communication line or the like, and information about threshold excess is automatically reported.
By performing remote maintenance, the reliability and maintainability of the information processing device can be improved.

[Brief description of drawings]

FIG. 1 is a block diagram showing an example of the configuration of an information processing system in which a failure information management system according to an embodiment of the present invention is implemented.

FIG. 2 is a block diagram showing an example of the principle of threshold management in the fault information management system of the present invention.

FIG. 3 is a flowchart showing an example of the operation of the fault information management system according to the exemplary embodiment of the present invention.

4A to 4G are conceptual diagrams showing an example of threshold value management in a failure information management method according to an embodiment of the present invention.

FIG. 5 is a conceptual diagram showing an example of an error code system and threshold management grouping in the fault information management system according to the embodiment of the present invention.

[Explanation of symbols]

1 ... Detailed error log collection part, 2 ... Error code creation part,
3 ... Error code creation table, 4 ... Error code storage unit, 5 ... Threshold management unit, 6 ... Fault information input / output unit, 7 ... Input / output device, 8 ... Detailed error log file, 9 ... Error code file, 10 ... Threshold value Management file, 10a ... threshold management table, 11 ... threshold information file, 12 ... communication control unit, 13 ... communication device, 14 ... maintenance center

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Yasuhiro Ishii 810 Shimoimaizumi, Ebina-shi, Kanagawa Hitachi Systems Office Systems Division

Claims (1)

[Claims] 1. A failure information management method for creating an error code corresponding to a failure when the failure occurs and managing the failure based on the error code and the occurrence frequency of the failure, wherein each error code is managed. Set the upper limit value and the lower limit value as a threshold value in a table, and when the failure occurs, obtain the upper limit value and the lower limit value corresponding to the error code of the failure from the table, and the number of failure occurrences is the lower limit value. As described above, the fault information management method is characterized in that the fault is judged to have exceeded the allowable range when the upper limit value or less.