JPH07141265A - Error monitoring method for magnetic disk device - Google Patents

Abstract

PURPOSE:To minimize manual operation in a monitoring process for error information by automating the short-time monitoring and long-period monitoring of the error information. CONSTITUTION:A short-period comparison part 7 inputs an error reference frequency from an error reference frequency parameter file 2 and compares data totalized by a short-period classification totalization part 6, and an error accumulating process part 8 stores the comparison result of the short-period comparison part 7 in a disk system error accumulation file 4. An information selection part 11 inputs conditions for selection from an information selection condition storage file 3 and selects necessary data in the result of totalization by a long-period classification totalization part 10, and a long-period comparison part 12 inputs an error reference frequency from the error reference frequency parameter file 2 and compares it with the data selected by the long-period classification totalization part 11. Then, an alarm output part 13 places an output device 14 according to the decision result of the short-period comparison part 7 or long-period comparison part 13 to generate an alarm.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an error monitoring method for monitoring the occurrence status of errors relating to a magnetic disk unit in an electronic computer system.

[0002]

2. Description of the Related Art FIG. 4 is a flow chart showing an example of a conventional error monitoring method for a magnetic disk device.

A conventional error monitoring method for monitoring the occurrence status of an error relating to a magnetic disk device in an electronic computer system is to output an error information report relating to an error of the magnetic disk device and maintain it (or an operator).
Is generally adopted, or a method of notifying a maintenance person (or an operator) of the occurrence of an error whenever it occurs is generally adopted.

Therefore, the maintenance personnel (or the operator) who has been informed of the error occurrence status by the error information report or the error notification notifies the magnetic disk device in which the error has occurred,
When the types of errors that have occurred are categorized and tabulated, and the signs of media failure of the magnetic disk drive are recognized from the results of the tabulation, preventive measures against failure are taken for the corresponding magnetic disk drive.

FIG. 4 is a flow chart showing an example of such a conventional error monitoring method for a magnetic disk device.

In FIG. 4, an error information storage file 1
Stores error information notified from each hardware of the electronic computer system to the operating system. The electronic computer system reads the error information in the error information storage file 1 (step 41), extracts only the information relating to the magnetic disk device therein (step 42), and edits it (step 43) for one day. An error information report is output once (step 44).

For example, when a particular "error event A" occurs three times in a particular magnetic disk device, and if this "error event A" occurs five times, the failure prevention for the magnetic disk device is prevented. If it is decided to take measures, it is judged from the above-mentioned error information report that "it is not necessary to take preventive measures against the occurrence of failure".

However, if the "error event A" has occurred even before the previous day, and the number of occurrences has not reached the reference value of 5, then this magnetic disk drive has a failure due to the "error event A". Despite the high probability of occurrence, the result is that "no preventive measures against the occurrence of the disorder will be taken."

As described above, since the error information before the previous day is not used for judging the necessity of preventive measures against the occurrence of the failure on the day, the error information before the previous day is used as a material for judging whether the preventive measures against the failure occurs. In order to be
Must review all error information for a certain period (for example, one week or one month) and check the history of occurrence of past error information.

[0010]

As described above, according to the conventional error monitoring method for the magnetic disk device, in order to monitor the occurrence status of the error event including the error information before the previous day, the maintenance personnel (or the operation person). Member) must review all the error information for a certain period of time and check the history of occurrence of past error information, so not only is it a great burden in terms of man-hours, but the work is performed visually. It has a drawback that it depends on the experience of a maintenance person (or an operator) and sometimes there is a risk of overlooking important error information and causing a failure of the magnetic disk device.

[0011]

According to the error monitoring method for a magnetic disk device of the present invention, only the error information about the magnetic disk device is stored in an error information storage file that stores information about errors in the hardware of the computer system. An error that is extracted, the error information is classified by device of the magnetic disk device and classified by error type to obtain a short-term aggregation result, and the short-term aggregation result is set as the upper limit number of times by the error type when the job is submitted. Compared with a reference count, when the short-term count result exceeds the error reference count, an alarm is output, and the short-term count result is accumulated and stored in a disk-based error cumulative file and stored in the disk-based error cumulative file. All the information that is stored is categorized by device of the magnetic disk device and by the type of error and then aggregated and lengthened. Obtaining the aggregation result, selecting only the information that matches the preset information selection condition from the long-term aggregation result to obtain the information selection result, and comparing the information selection result with the error reference count to obtain the information. It includes outputting an alarm when the selection result exceeds the error reference number, and further incorporates it into the job automatic operation system to automatically operate at a constant cycle. Is a generation management file having files of a plurality of generations that are new at regular intervals, and the constant period can be set to an arbitrary length.

[0012]

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

FIG. 1 is a flow chart showing an embodiment of the present invention, FIG. 2 is a block diagram expressing the embodiment of FIG. 1 as a functional block, and FIG. 3 shows an example of contents of each file of the embodiment of FIG. In the format diagram, (a) is an error information storage file, (b) is an error reference parameter file,
(C) is a diagram showing a disk-based error accumulation file, and (d) is an information selection condition file.

In this embodiment, as shown in FIG. 2, an error information storage file 1 for storing information about errors in the hardware of the electronic computer system and error contents for each type of error in the magnetic disk unit. Error reference count parameter file 2 in which the upper limit count determined from the risk is set as the reference count for the error and stored, and the disk error cumulative file that accumulates and stores the error information of the magnetic disk device 4, an information selection condition storage file 3 that stores conditions for selecting necessary information from the error information stored in the disk-based error accumulation file 4, and an output device 14 that outputs an alarm. The error information input / extraction unit 5 inputs the error information from the error information storage file 1 to input the error information to the magnetic disk device. Error information extracted by the error information input / extraction unit 5 is categorized by the error type for each magnetic disk device and aggregated by the short-term comparison unit 7. In, the error reference number is input from the error reference number parameter file 2, the data aggregated by the short-term classification and aggregation unit 6 is compared, and the error accumulation processing unit 8 compares the result by the short-term comparison unit 7 with the disk error. The information stored in the cumulative file 4 is input to the cumulative error input unit 9, and the information stored in the disk error cumulative file 4 is input to the cumulative error input unit 9. Each device is classified by type of error and totaled, and the information selection unit 11 inputs the conditions for selection from the information selection condition storage file 3. Then, necessary data is selected from the results aggregated by the long-term classification aggregation unit 10 according to the conditions, and the long-term comparison unit 12 inputs the error reference number from the error reference number parameter file 2 to perform long-term classification aggregation. In comparison with the data selected by the unit 11, the alarm output unit 13 operates the output device 14 according to the determination result of the short-term comparison unit 7 or the long-term comparison unit 12 to generate an alarm.

The above operation will be described in detail below with reference to FIGS. 1 and 3.

All the error information of the hardware of the electronic computer system including the magnetic disk device is stored in the error information storage file 1 by the operating system.

The error information storage file 1 is shown in FIG.
As shown in FIG. 3, the error occurrence time 51, the device name 52 in which the error occurred, and the content of the error (error content) 53 are recorded as a set.

In step 21 of FIG. 1, the above error information is input from the error information storage file 1,
In step 22, only the error information regarding the magnetic disk device is extracted from them. Subsequently, in step 23, the error information regarding the magnetic disk device extracted in step 22 is classified by the type of error, and in step 24, the classified error information is totaled for each magnetic disk device.

Next, at step 25, the error reference number set for each type of error is input from the error reference number parameter file 2. As shown in FIG. 3B, the error reference number parameter file 2 includes an error type 61 in which each error is indicated by a symbol, a reference number 62 set for each error, and the content of each error. The described error type description 63 is recorded as a set.

Next, in step 26, step 24
The number of error information for each magnetic disk device
The corresponding reference number of the error reference number parameter file 2 is compared, and when the number of error information exceeds the reference number, the alarm output unit is notified of the device number of the magnetic disk device, the error count, and the error type, The alarm output unit causes the output device 14 to output a message (step 3).
5).

Regardless of the comparison result between the error information and the reference number in step 25, the totalized result in step 24 is stored in the disk-based error cumulative file 4 (step 27).

The disk system error cumulative file 4 is shown in FIG.
As shown in (c), for all errors that occurred in the past in the magnetic disk device, the device name 7 in which the error occurred
1, an error occurrence time 72, an error type 73 in which the error type is indicated by a symbol, and an error content (error content) 74 are randomly recorded as one set.

Next, in step 28, the error information accumulated from the disk-type error accumulation file 4 is input, in step 29 the error information is classified according to the type of error, and in step 30, they are totaled.

Since the totalized result in step 30 includes unnecessary errors that should be excluded from the judgment target because the treatment has already been completed, in step 31,
The conditions for selecting only necessary errors are input from the information selection condition storage file 3.

The information selection condition storage file 3 is shown in FIG.
As shown in (d), the device name 81, the selection key date and time 82 indicating the date and time when the device is treated, and the comment 83 indicating the content of the action performed on the date and time are recorded as a set. Has been done. Therefore, from this information, the next object to be processed is selected (step 32).

Next, at step 33, the error reference number is input from the error reference number parameter file 2,
In step 34, the number of error information for each magnetic disk device tabulated in step 30 is compared with the corresponding error reference number. If the number of times of error information exceeds the reference number, the alarm output unit is notified of the device number of the magnetic disk device, the error count, and the error type, and the alarm output unit causes the output device 14 to output a message (step 3).
5). If the number of error information does not exceed the reference number,
The process is terminated as it is.

Thus, in addition to monitoring the error information at regular intervals, a process of automatically comparing the number of occurrences of error information over a long period of time with a reference number and monitoring it can be executed as a series of shoves. By doing so, human intervention in the error information monitoring process can be suppressed to a minimum. Also, since the monitoring accuracy can be improved,
It is possible to prevent a failure of the magnetic disk device from occurring.

It should be noted that, although the above description is for one-time processing operation, by incorporating this operation in the job automatic operation system, the error information monitoring processing can be automatically executed at a constant cycle. You can

Since the content of the error reference number parameter file can be changed when the job is submitted, the alarm generation level can be arbitrarily set according to the characteristics of the magnetic disk device.

Further, the disk-based error cumulative file is made new at regular intervals, and a generation management file (for example, one disk-based error cumulative file has a cumulative period of one month) by a plurality of disk-based error cumulative files of different periods. , When the data for 6 months is targeted,
It is possible to arbitrarily set the length of the cumulative period by configuring the generation management file by the six disk-type error cumulative files).

[0031]

As described above, the error monitoring method of the magnetic disk device according to the present invention includes short-term monitoring for monitoring the result of independently collecting error information generated in the magnetic disk device at regular intervals. By automatically enabling long-term monitoring that monitors the result of totaling error information that has occurred in the cumulative period set to an arbitrary length, human intervention for error information monitoring processing can be performed. There is an effect that it can be suppressed to the minimum. Further, since the monitoring accuracy can be improved, it is possible to prevent occurrence of a failure in the magnetic disk device.

[Brief description of drawings]

FIG. 1 is a flowchart showing an embodiment of the present invention.

FIG. 2 is a block diagram showing the embodiment of FIG. 1 as a functional block.

3 is a format diagram showing an example of the contents of each file of the embodiment of FIG. 1, (a) is an error information storage file,
(B) is a diagram showing an error reference parameter file, (c) is a disk-based error cumulative file, and (d) is an information selection condition file.

FIG. 4 is a flowchart showing an example of a conventional error monitoring method for a magnetic disk device.

[Explanation of symbols]

 1 Error information storage file 2 Error reference count parameter file 3 Information selection condition storage file 4 Disk error accumulation file 5 Error information input / extraction unit 6 Short period classification aggregation unit 7 Short period comparison unit 8 Error accumulation processing unit 9 Accumulated error input Part 10 Long-term classification / aggregation part 11 Information selection part 12 Long-term comparison part 13 Alarm output part 14 Output device 21-35.41-43 Step 51.72 Error occurrence time 52.71.81 Device name 53.74 Error content 61/73 Error type 62 Reference count 63 Error type description 82 Selection key date / time 83 Comment

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI technical display location G11B 20/18 574 E 9074-5D

Claims (3)

[Claims] 1. An error information storage file that stores information related to general hardware errors of an electronic computer system is used to extract only error information related to a magnetic disk device, and the error information is used for each device of the magnetic disk device and the error information. The short-term aggregation result is obtained by classifying and summing up by type, and the short-term aggregation result is compared with the error reference number that is set as the upper limit number for each error type when the job is submitted. When it exceeds the limit, an alarm is output, the short-term totaling result is accumulated and stored in the disk-based error cumulative file, and all the information stored in the disk-based error cumulative file is stored for each device of the magnetic disk device and the error Calculate the long-term aggregation results by classifying by type and set in advance from the long-term aggregation results. To obtain an information selection result by selecting only information that matches a certain information selection condition, compare the information selection result with the error reference number, and issue an alarm when the information selection result exceeds the error reference number. An error monitoring method for a magnetic disk device, comprising: 2. A method for monitoring an error in a magnetic disk drive according to claim 1, wherein the method is incorporated into an automatic job operation system and automatically operated at a constant cycle. 3. A generation management file having files of a plurality of generations, in which the disk-type error cumulative file is made new every fixed period, and the fixed period can be set to an arbitrary length. An error monitoring method for a magnetic disk device according to claim 1 or 2.