JP2011258261A - Deterioration level determination device, computer program and deterioration level determination method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To determine deterioration level of a device to be diagnosed based on measured values even when the relationship between the deterioration of the device to be diagnosed and measured values is unclear.SOLUTION: A normal sample value storage 6 (normal value storage) stores a normal sample value (diagnosis information normal value). A diagnosis information collection section 5 (measured value obtaining section) collects diagnosis information (diagnosis information measured value). A diagnosis information difference calculation section 7 (difference calculation section) calculates diagnosis information difference (index difference) based on the diagnosis information and the normal sample value. A disc drive deterioration level determination section 8 (deterioration level determination section) determines the deterioration level based on the diagnosis information difference.

Description

  The present invention relates to a degradation level determination device that determines a degradation level of a diagnosis target device.

  There is a method of determining the degree of deterioration of the diagnosis target device by observing the diagnosis target device, acquiring the observation value, and comparing the acquired observation value with a threshold value.

JP 2008-148226 A JP 2009-266291 A

In order to determine the degree of deterioration by comparing the observed value with the threshold value, it is necessary to set the threshold value in advance. In order to set the threshold value in advance, it is necessary to know in advance how much the deterioration of the diagnosis target device has advanced and how the observed value changes.
The present invention has been made to solve the above-described problem, for example, and even if it is not known in advance how the observed value changes when the diagnosis target device progresses, the observed value It is an object of the present invention to be able to determine the degree of deterioration of a diagnosis target device based on the above.

The degradation degree determination device according to the present invention is:
A processing device that processes data, a storage device that stores data, an observation value acquisition unit, a normal value storage unit, a distance calculation unit, and a deterioration degree determination unit;
The observation value acquisition unit acquires, as the diagnostic information observation value, a numerical value that changes as the diagnosis target device deteriorates, using the processing device.
The normal value storage unit uses the storage device to store a plurality of diagnostic information observation values as diagnostic information normal values in a state where the diagnosis target device is not deteriorated,
The distance calculation unit uses the processing device to perform a plurality of diagnosis based on the diagnosis information observation value acquired by the observation value acquisition unit and the plurality of diagnosis information normal values stored by the normal value storage unit. A numerical value indicating how far the diagnostic information observation value is from the set of normal information values is calculated as an index distance,
The deterioration degree determination unit is configured to determine a deterioration degree of the diagnosis target device using the processing device based on the index distance calculated by the distance calculation unit.

  According to the degradation degree determination device according to the present invention, even if it is not known in advance how the diagnostic information observation value changes when the degradation of the diagnostic target device progresses, the diagnostic target is based on the diagnostic information observation value. The degree of deterioration of the device can be determined.

1 is a configuration diagram showing a disk drive deterioration degree determination device 2 in Embodiment 1. FIG. The figure which shows an example in case the distance with a normal sample value is small. The figure which shows the example in case the distance with a normal sample value is separated a little. FIG. 4 is a configuration diagram showing a disk drive deterioration degree determination device 2 according to a second embodiment. FIG. 6 is a configuration diagram showing a disk drive deterioration degree determination device 2 according to a third embodiment. FIG. 10 is a perspective view illustrating an example of an appearance of a deterioration degree determination device 100 according to a fourth embodiment. FIG. 10 is a diagram illustrating an example of hardware resources of a degradation degree determination apparatus 100 according to Embodiment 4. FIG. 10 is a system configuration diagram illustrating an example of an overall configuration of a deterioration degree determination system 800 according to a fourth embodiment. The flowchart figure which shows an example of the flow of deterioration degree determination processing S510 in Embodiment 4. FIG. 10 is a system configuration diagram illustrating an example of an overall configuration of a deterioration degree determination system 800 according to a fifth embodiment. FIG. 20 is a flowchart showing an example of a flow of promotion processing S520 in the fifth embodiment.

In maintenance and inspection work for social infrastructure equipment such as power-related equipment and air traffic control equipment, preventive maintenance represented by deterioration diagnosis before failure is required in addition to periodic inspection and failure repair. In addition, in the power information system, there is a demand for replacement from a mainframe to a general-purpose server, but the conventional high reliability is also required.
As a method for realizing high reliability of a computer, failure detection of a hard disk drive (hereinafter abbreviated as HDD) is performed. Recent HDDs include S.I. M.M. A. R. T.A. A function called (Self-Monitoring Analysis and Reporting Technology) is installed, and various information (SMART value) such as the number of replaced defective sectors and the number of spin retries can be acquired. In the HDD deterioration diagnosis, the collected S.D. M.M. A. R. T.A. A method based on comparison between a value and a set threshold value is used. For example, the S.D. M.M. A. R. T.A. When the value exceeds the threshold value, data is protected by restricting HDD access. There is also a method in which a deterioration model is generated instead of a threshold value, and deterioration determination is performed by comparison with the model. For example, based on the sensor item of the disk (SMAT R.T.) and the maximum current required to reach the specified rotational speed measured using a special device, the sensor item and the spindle of the disk A deterioration estimation model that expresses the relationship between the deterioration level of the motor lubricating oil and a HDD survival model that expresses the relationship between the change in the time series of the deterioration degree of the lubricating oil and the survival probability of the disk were generated and acquired from the diagnosis target disk. The sensor information is applied to this model to predict the failure time of the diagnosis target disk.

  In the diagnosis method based on comparison with the threshold value, there is a possibility that it is erroneously judged as deterioration when the threshold value is temporarily exceeded (instantly), and whether the deterioration is mild or severe (failure is close). Cannot be determined. In addition, the diagnostic method based on comparison with a model is poor in versatility because it is necessary to construct a model for each vendor and each disk type. In addition, a special device for constructing a model is required, which is expensive.

  The degradation level determination apparatus described below realizes a highly accurate HDD degradation level determination method and enables appropriate processing according to the degree of degradation (light degradation / heavy degradation). In addition, the degradation level determination apparatus does not require model construction and realizes a non-vendor-dependent and low-cost degradation level determination method.

Embodiment 1 FIG.
The first embodiment will be described with reference to FIGS.

FIG. 1 is a configuration diagram showing a disk drive deterioration degree judging device 2 in this embodiment.
The disk drive deterioration degree determination apparatus 2 (deterioration degree determination apparatus) is a disk drive deterioration degree determination apparatus built in or connected to the computer 1. The operating system 3 is an operating system of the computer 1. The disk drive 4 (diagnosis target device) is one or more disk drives built in or connected to the computer 1. The diagnostic information collection unit 5 (observation value acquisition unit) is a diagnostic information collection unit that collects diagnostic information from the disk drive 4 via the operating system 3 at a predetermined period. The normal sample value storage unit 6 (normal value storage unit) is a normal sample value storage unit that stores diagnostic information collected from a normal disk in advance. The diagnosis information distance calculation unit 7 (distance calculation unit) calculates the distance between the diagnosis information collected by the diagnosis information collection unit 5 and the diagnosis information of the normal disk stored in the normal sample value storage unit 6 by a statistical method. It is an information distance calculation unit. The disk drive degradation level determination unit 8 (degradation level determination unit) is a disk drive degradation level determination unit that determines the degradation level of the disk drive based on the distance calculated by the diagnostic information distance calculation unit 7. The disk drive degradation level notification unit 9 (degradation level notification unit) is a disk drive degradation level notification unit that notifies the outside of the computer of the degradation level of the disk drive determined by the disk drive degradation level determination unit 8.

Next, the operation will be described.
(1) Collect diagnostic information (one or more types of information related to disk drive diagnosis that can be collected by SMART or the like) from one or more normal disks and store them in the normal sample value storage unit 6 in advance.
(2) The diagnostic information collection unit 5 collects one or more types of diagnostic information from the disk drive 4 via the operating system 3 at a predetermined cycle (for example, once an hour).
(3) The distance between the diagnostic information of the normal disk stored in advance in (1) and the diagnostic information collected in (2) is calculated by a statistical method by the diagnostic information distance calculation unit 7. Statistical methods used to calculate the distance include a Wilcoxon rank sum test and a support vector machine (Support Vector Machine).

FIG. 2 is a diagram illustrating an example when the distance between the normal sample value and the diagnostic information is small.
Open circles indicate normal sample values. The diamonds are the collected S.P. M.M. A. R. T information is shown.
In this case, since the distance is small, it is normal.

FIG. 3 is a diagram illustrating an example when the distance between the normal sample value and the diagnostic information is slightly separated.
In this case, since the distance is slightly apart, the deterioration is slight.

(4) The disk drive deterioration degree determination unit 8 classifies the deterioration degree of the disk drive according to the distance calculated in (3). For example, it is classified into three stages of normal, light deterioration, and heavy deterioration (normal when the distance is small, and deterioration when the distance is large).
(5) The disk drive deterioration degree notification unit 9 notifies the outside of the computer of the disk drive deterioration degree determined in (4).

  With such a configuration and operation, the degree of deterioration of the disk drive can be accurately and finely determined, so that it is possible to appropriately perform measures such as data backup and disk drive replacement.

  When the Wilcoxon rank sum test is used to calculate the distance of (3), an index called p-value is calculated for two sets (normal sample value and diagnosis information of the diagnosis target disk), and the smaller this value, the more 2 It is determined that the divergence degree of the two sets is high (that is, the distance is large). In the determination of the degree of deterioration in (4), for example, when the p value is smaller than 0.05, the heavy deterioration is performed, when the p value is smaller than 0.25, the light deterioration is performed, and the other cases are determined to be normal. Can be classified into stages. Moreover, it can also classify | categorize more finely based on p value.

加重平均は、例えば、Ｓ．Ｍ．Ａ．Ｒ．Ｔ．のＩＤ１（Ｒａｗ Ｒｅａｄ Ｅｒｒｏｒ Ｒａｔｅ）のｐ値をｐ_１、ＩＤ５（Ｒｅａｌｌｏｃａｔｅｄ Ｓｅｃｔｏｒｓ Ｃｏｕｎｔ）のｐ値をｐ_５、ＩＤ１９６（Ｒｅａｌｌｏｃａｔｉｏｎ Ｅｖｅｎｔ Ｃｏｕｎｔ）のｐ値をｐ_１９６としたとき、ＩＤ５の寄与が他の２つより大きい場合に平均を（０．２×ｐ_１＋０．６×ｐ_５＋０．２×ｐ_１９６）÷（０．２＋０．６＋０．２）で求めるなどである。 When the distance is obtained from a plurality of types of diagnostic information, the p value is obtained for each piece of information, and the distance is obtained by averaging them. The calculation method of the average may be a simple arithmetic average or geometric average, or a weighted average obtained by weighting according to the contribution to the degree of deterioration of each diagnosis information. Shows the each formula below (weighted w _i in average weight).

The weighted average is, for example, S.I. M.M. A. R. T.A. Of ID1 (Raw Read Error Rate) p value _p 1 of, ID5 (Reallocated Sectors Count) p value _p 5 of, ID196 when the p-value (Reallocation Event Count) and _{p 196,} the contribution of other ID5 If the number is larger than two, the average is calculated by (0.2 × p ₁ + 0.6 × p ₅ + 0.2 × p ₁₉₆ ) ÷ (0.2 + 0.6 + 0.2).

この場合、正常標本値と診断情報との差の二乗の和に基づいて距離を算出しているため、正常標本値と診断情報との差が二乗のオーダーで距離の大きさに寄与することになる。このため、正常標本値との差が大きくなるほど急激に劣化が進む診断情報の影響を表すことができる。 In the above description, the method using the Wilcoxon rank sum test is shown as the distance calculation method, but any evaluation function may be used. For example, assuming that the average of normal sample values N _x (x = 1 to j) is N and the average of diagnostic information T _y (y = 1 to k) of the target disk is T, the distance D is defined by the following equation: Is possible.

In this case, since the distance is calculated based on the sum of the squares of the difference between the normal sample value and the diagnostic information, the difference between the normal sample value and the diagnostic information contributes to the magnitude of the distance in the order of the square. Become. For this reason, it is possible to represent the influence of diagnostic information whose deterioration rapidly increases as the difference from the normal sample value increases.

  The disk drive deterioration level determination device 2 described above includes a diagnostic information collection unit 5, a normal sample value storage unit 6, a diagnostic information distance calculation unit 7, and a disk drive deterioration level determination unit 8. The diagnostic information collection unit 5 collects one or more types of diagnostic information designated from a disk drive having a function capable of acquiring one or more types of diagnostic information by transmitting a command at a predetermined cycle. The normal sample value storage unit 6 stores diagnostic information acquired from a disk drive that normally operates in advance. The diagnostic information distance calculation unit 7 calculates the distance between the diagnostic information acquired by the diagnostic information collection unit 5 and the diagnostic information stored in the normal sample value storage unit 6 by a statistical method. The disk drive deterioration degree determination unit 8 determines the deterioration degree of the disk drive based on the distance calculated by the diagnostic information distance calculation unit 7, and classifies the deterioration degree of the disk drive into two or more types.

  Further, the disk drive deterioration degree determination device 2 includes a disk drive deterioration degree notification unit 9. The disk drive deterioration degree notification unit 9 notifies the disk drive deterioration degree determined by the disk drive deterioration degree determination unit 8 to the outside.

  In addition, the disk drive deterioration degree determination device 2 uses S.D. M.M. A. R. T.A. (Self-Monitoring Analysis and Reporting Technology) is used.

Embodiment 2. FIG.
Embodiment 2 will be described with reference to FIG.
In addition, about the part which is common in Embodiment 1, the same code | symbol is attached | subjected and description is abbreviate | omitted.

FIG. 4 is a configuration diagram showing the disk drive deterioration degree judging device 2 in this embodiment.
The diagnostic information collection cycle calculation unit 10 is a diagnostic information collection cycle calculation unit that calculates a collection cycle by the diagnostic information collection unit 5 based on the deterioration degree determined by the disk drive deterioration degree determination unit 8. The collected diagnostic information changing unit 11 is a collected diagnostic information changing unit that changes the items of diagnostic information collected by the diagnostic information collecting unit 5 based on the degree of deterioration determined by the disk drive deterioration degree determining unit 8. The degradation level notification method change unit 12 is a degradation level notification method change unit that changes the notification method by the disk drive degradation level notification unit 9 based on the degradation level determined by the disk drive degradation level determination unit 8. Others are the same as in the first embodiment.
Next, the operation will be described with respect to differences from the first embodiment.
(1) After the operation (4) of the first embodiment, the diagnostic information collection cycle calculation unit 10 appropriately calculates the diagnostic information collection cycle based on the deterioration degree determined by the disk drive deterioration degree determination unit 8. To do. For example, the diagnostic information is collected at a period of 70% at the normal time in the case of light deterioration, and at a short period of 10% at the normal time in the case of heavy deterioration.
The collection period of diagnostic information can be the same period for all diagnostic information, or can be set individually for each diagnostic information. For example, when it is determined that the deterioration is minor, the diagnostic information is collected at a period of 50% of normal for diagnostic information having the longest distance among the diagnostic information and at the same period as for normal for other diagnostic information. You can also.
(2) The collected diagnostic information changing unit 11 appropriately changes the type of diagnostic information to be collected based on the degree of deterioration determined by the disk drive deterioration degree determining unit 8. For example, three types of information are collected when normal, seven types of information are collected in addition to the three types of normal when lightly deteriorated, and all types of information are collected when severely deteriorated.
(3) The deterioration degree notification method change unit 12 appropriately changes the deterioration degree notification method based on the deterioration degree determined by the disk drive deterioration degree determination unit 8. For example, in the case of light deterioration, notification is made by e-mail, and in the case of heavy deterioration, notification is also made by sounding an alarm sound and blinking of a warning light together with e-mail notification.

  The disk drive deterioration degree determination apparatus 2 described above includes a diagnostic information collection cycle calculation unit 10. The diagnostic information collection cycle calculation unit 10 calculates the collection cycle by the diagnostic information collection unit 5 based on the disk drive deterioration level determined by the disk drive deterioration level determination unit 8.

  Further, the disk drive deterioration degree determination device 2 includes a collected diagnosis information change unit 11. The collected diagnostic information changing unit 11 changes the items of diagnostic information collected by the diagnostic information collecting unit 5 based on the disk drive deterioration level determined by the disk drive deterioration level determination unit 8.

  Further, the disk drive deterioration degree determination device 2 includes a deterioration degree notification method changing unit 12. The deterioration level notification method changing unit 12 changes the notification method by the disk drive deterioration level notification unit 9 based on the disk drive deterioration level determined by the disk drive deterioration level determination unit 8.

  With such a configuration and operation, when the deterioration progresses by shortening the collection cycle at the time of deterioration, it becomes possible to quickly detect it. If the period is individually set for each diagnosis information, the progress of the deterioration can be detected more quickly by collecting the diagnosis information whose deterioration is progressing in a shorter period. In addition, it collects information on a small number of items during normal operation and increases the types of information collected during deterioration, thereby reducing the load of collecting diagnostic information during normal operation. Judgment can be made. In addition, by changing the notification method according to the degree of deterioration, in the case of minor deterioration that is not high in urgency, it is possible to reduce the inconvenience of the user by notifying by e-mail, and in the case of heavy deterioration where the failure is near and the urgency is high Can be promptly notified by an alarm sound or warning light.

Embodiment 3 FIG.
Embodiment 3 will be described with reference to FIG.
In addition, about the part which is common in Embodiment 1- Embodiment 2, the same code | symbol is attached | subjected and description is abbreviate | omitted.

FIG. 5 is a configuration diagram showing the disk drive deterioration degree judging device 2 in this embodiment.
The diagnostic information and degradation level determination result storage unit 13 stores the diagnostic information collected by the diagnostic information collection unit 5 and the history of the disk drive degradation level determined by the disk drive degradation level determination unit 8. A determination result storage unit. Diagnosis information distance calculation method and deterioration degree determination method The dynamic change unit 14 determines the diagnosis information and deterioration degree determination method by calculating the distance calculation method in the diagnosis information distance calculation unit 7 and the deterioration degree determination method in the disk drive deterioration degree determination unit 8. It is a diagnostic information distance calculation method and a deterioration degree determination method dynamic change section that change based on the diagnosis information stored in the result storage section 13 and the deterioration degree determination result. Others are the same as in the first embodiment.

Next, the operation will be described with respect to differences from the first embodiment.
(1) After the operation (4) or (5) of the first embodiment (after collecting diagnostic information, calculating the distance, and determining the disk deterioration level), the date and time when the deterioration level determined as the collected diagnostic information is collected Together with the information, it is stored in the diagnostic information and deterioration level determination result storage unit 13. Hereinafter, such information is referred to as history information.
(2) Diagnostic Information Distance Calculation Method and Degradation Level Determination Method The dynamic change unit 14 acquires history information from the diagnostic information and degradation level determination result storage unit 13. The acquisition is performed every time a predetermined period or the amount of history information stored in the diagnostic information and deterioration degree determination result storage unit 13 exceeds a predetermined size.
(3) Diagnostic Information Distance Calculation Method and Degradation Level Determination Method The dynamic change unit 14 determines the distance calculation method in the diagnostic information distance calculation unit 7 and the deterioration in the disk drive deterioration level determination unit 8 based on the acquired history information. Change the degree judgment method.
For example, in the history information determined to be light deterioration or heavy deterioration, when diagnosis information with a large distance is concentrated on a specific type of diagnosis information (for example, the distance of the specific type of diagnosis information is larger than the diagnosis information of another type) If the weight of the specific type of diagnostic information is increased, the weight of the weighted average is set (for example, the weight of the diagnostic information of the specific type is 0.5 and the total weight of other diagnostic information). The distance calculation method is changed so that the distance is calculated. As a result, it is possible to determine the degree of deterioration reflecting the influence of diagnostic information that is considered to be particularly prone to deterioration.
Also, if there is too much history information determined to be heavy deterioration than history information determined to be light deterioration (for example, heavy deterioration is more than twice the light deterioration), it is considered that the determination method of the degree of deterioration should be improved. As for p value in Wilcoxon rank sum test, if the p value is less than 0.10, it is classified as heavy deterioration, and if the p value is less than 0.20, it is classified as light deterioration. By lowering the threshold value of 0.05 by 0.05 and increasing the light degradation p-value threshold value by 0.05 to 0.25, more appropriate classification becomes possible.

  The disk drive degradation level determination device 2 described above includes a diagnostic information and degradation level determination result storage unit 13 and a diagnostic information distance calculation method and degradation level determination method dynamic change unit 14. The diagnostic information and deterioration level determination result storage unit 13 stores collected diagnostic information and a history of deterioration level determination. Diagnosis information distance calculation method and deterioration degree determination method The dynamic change unit 14 determines the diagnosis information and deterioration degree determination method by calculating the distance calculation method in the diagnosis information distance calculation unit 7 and the deterioration degree determination method in the disk drive deterioration degree determination unit 8. It changes based on the diagnostic information stored in the result storage unit 13 and the deterioration degree determination result.

  This configuration / operation enables feedback to appropriately change the distance calculation method of the diagnostic information and the determination method of the degree of deterioration of the disk drive based on the history information, improving the detection accuracy of deterioration and rapid deterioration. Can be detected.

Embodiment 4 FIG.
The fourth embodiment will be described with reference to FIGS.
In addition, about the part which is common in Embodiment 1- Embodiment 3, the same code | symbol is attached | subjected and description is abbreviate | omitted.

FIG. 6 is a perspective view showing an example of the appearance of the degradation degree determination apparatus 100 in this embodiment.
The degradation degree determination apparatus 100 includes a system unit 910, a display device 901 having a CRT (Cathode / Ray / Tube) or LCD (liquid crystal) display screen, a keyboard 902 (Key / Board: K / B), a mouse 903, an FDD 904 ( (Flexible / Disk / Drive), compact disk device 905 (CDD), printer device 906, scanner device 907, and other hardware resources, which are connected by cables and signal lines.
The system unit 910 is a computer, and is connected to the facsimile machine 932 and the telephone 931 via a cable, and is connected to the Internet 940 via a local area network 942 (LAN) and a gateway 941.

FIG. 7 is a diagram illustrating an example of hardware resources of the degradation degree determination apparatus 100 according to this embodiment.
The degradation degree determination apparatus 100 includes a CPU 911 (also referred to as a central processing unit, a central processing unit, a processing unit, an arithmetic unit, a microprocessor, a microcomputer, and a processor) that executes a program. The CPU 911 is connected to the ROM 913, the RAM 914, the communication device 915, the display device 901, the keyboard 902, the mouse 903, the FDD 904, the CDD 905, the printer device 906, the scanner device 907, and the magnetic disk device 920 via the bus 912, and the hardware. Control the device. Instead of the magnetic disk device 920, a storage device such as an optical disk device or a memory card read / write device may be used.
The RAM 914 is an example of a volatile memory. The storage media of the ROM 913, the FDD 904, the CDD 905, and the magnetic disk device 920 are an example of a nonvolatile memory. These are examples of a storage device or a storage unit. A communication device 915, a keyboard 902, a scanner device 907, an FDD 904, and the like are examples of an input unit and an input device.
Further, the communication device 915, the display device 901, the printer device 906, and the like are examples of an output unit and an output device.

The communication device 915 is connected to a facsimile machine 932, a telephone 931, a LAN 942, and the like. The communication device 915 is not limited to the LAN 942, and may be connected to the Internet 940, a WAN (wide area network) such as ISDN, or the like. When connected to a WAN such as the Internet 940 or ISDN, the gateway 941 is unnecessary.
The magnetic disk device 920 stores an operating system 921 (OS), a window system 922, a program group 923, and a file group 924. The programs in the program group 923 are executed by the CPU 911, the operating system 921, and the window system 922.

The program group 923 stores programs that execute functions described as “˜units” in the description of the embodiments described below. The program is read and executed by the CPU 911.
In the file group 924, information, data, signal values, variable values, and parameters described as “determination results of”, “calculation results of”, and “processing results of” in the description to be described below are “ It is stored as each item of “file” and “˜database”. The “˜file” and “˜database” are stored in a recording medium such as a disk or a memory. Information, data, signal values, variable values, and parameters stored in a storage medium such as a disk or memory are read out to the main memory or cache memory by the CPU 911 via a read / write circuit, and extracted, searched, referenced, compared, and calculated. Used for CPU operations such as calculation, processing, output, printing, and display. Information, data, signal values, variable values, and parameters are temporarily stored in the main memory, cache memory, and buffer memory during the CPU operations of extraction, search, reference, comparison, operation, calculation, processing, output, printing, and display. Is remembered.
In addition, the arrows in the flowcharts described in the following description of the embodiments mainly indicate input / output of data and signals. The data and signal values are the RAM 914 memory, the FDD 904 flexible disk, the CDD 905 compact disk, and the magnetic field. The data is recorded on a recording medium such as a magnetic disk of the disk device 920, another optical disk, a mini disk, and a DVD (Digital Versatile Disk). Data and signals are transmitted online via a bus 912, signal lines, cables, or other transmission media.

  In the following description, what is described as “-unit” may be “-circuit”, “-apparatus”, “-apparatus”, and “-step”, “-procedure”, “ ~ Process ". That is, what is described as “˜unit” may be realized by firmware stored in the ROM 913. Alternatively, it may be implemented only by software, or only by hardware such as elements, devices, substrates, and wirings, by a combination of software and hardware, or by a combination of firmware. Firmware and software are stored as programs in a recording medium such as a magnetic disk, a flexible disk, an optical disk, a compact disk, a mini disk, and a DVD. The program is read by the CPU 911 and executed by the CPU 911. That is, the program causes the computer to function as “to part” described below. Alternatively, the procedure or method of “to part” described below is executed by a computer.

FIG. 8 is a system configuration diagram showing an example of the overall configuration of the deterioration degree determination system 800 in this embodiment.
The degradation level determination system 800 includes a diagnosis target device 810, an observation device 820, and a degradation level determination device 100.
The diagnosis target device 810 is a target device for determining the degree of deterioration. The diagnosis target device 810 is, for example, a disk drive device. The diagnosis target device 810 may be a device such as the magnetic disk device 920 built in the degradation level determination device 100 or may be a device that exists separately from the degradation level determination device 100.
The observation device 820 observes the diagnosis target device 810 and generates a numerical value representing the observation result. A numerical value generated by the observation device 820 is referred to as a “diagnosis information observation value”. The diagnostic information observation value changes as the diagnostic target device 810 deteriorates. The observation device 820 may be a device that observes the diagnosis target device 810 from the outside, or may be a device that is built in the diagnosis target device 810. The diagnosis information observation value is, for example, a SMART value. There may be one type of diagnostic information observation value or a plurality of types. The observation device 820 repeatedly observes the diagnosis target device 810 at a predetermined cycle or at a cycle designated by the degradation degree determination device 100, and generates a diagnostic information observation value. The diagnosis target device 810 outputs the generated diagnostic information observation value.

  The degradation level determination apparatus 100 determines the degradation level of the diagnosis target apparatus 810 based on the diagnostic information observation value generated by the observation apparatus 820. The degradation level determination apparatus 100 includes an observed value acquisition unit 110, an observed value storage unit 120, a normal value storage unit 130, a distance calculation unit 140, a degradation level determination unit 150, a degradation level storage unit 160, and a degradation level. And a notification unit 170.

The normal value storage unit 130 stores a plurality of diagnostic information observation values in a state where the diagnosis target device 810 is not deteriorated in advance using the magnetic disk device 920. The diagnostic information observation value stored in the normal value storage unit 130 is referred to as “diagnostic information normal value”.
For example, when the SMART value is used as the diagnostic information observation value, the SMART value calculation method varies depending on the manufacturer of the disk drive device. Also, the normal value may be different depending on the model of the disk drive device. Thus, although it is known that the diagnostic information observation value changes with the deterioration of the diagnosis target device 810, it can be said that the value is judged to be normal and the deterioration is advanced when the value is changed. However, a value that is not always obvious may be used. Further, there are cases where there is a correlation between a plurality of types of diagnostic information observation values, and it can be determined whether or not deterioration has progressed by the combination thereof.
Therefore, for example, a large number of diagnosis target devices 810 of the same model as the diagnosis target device 810 are observed, and a large amount of normal values of diagnosis information is collected in advance. The normal value storage unit 130 stores the diagnostic information normal values collected in this way in advance.

The observation value acquisition unit 110 uses the CPU 911 to acquire the diagnostic information observation value output by the observation device 820.
The observation value acquisition unit 110 may be configured to passively acquire the diagnostic item observation values that the observation device 820 outputs spontaneously. Alternatively, when the observation value acquisition unit 110 wants to acquire the diagnostic item observation value, the observation value acquisition unit 110 transmits a command specifying the observation item to the observation device 820, and the diagnosis output by the observation device 820 as a response to the transmitted command. The structure which acquires an item observation value may be sufficient. Alternatively, the observation value acquisition unit 110 has a configuration in which a command that specifies an observation period and an observation item is transmitted in advance, and a diagnostic item observation value that is periodically output by the observation device 820 is acquired according to the transmitted command. There may be.

The observation value storage unit 120 stores the diagnostic information observation values acquired by the observation value acquisition unit 110 using the magnetic disk device 920.
The observation value storage unit 120 may be configured to store only the diagnostic information observation value acquired last by the observation value acquisition unit 110. Alternatively, the observation value storage unit 120 may be configured to store a plurality of diagnostic information observation values acquired by the observation value acquisition unit 110 within a predetermined period. If it is determined whether or not the diagnosis target device 810 has been deteriorated only by one diagnosis information observation value, it is impossible to deal with an abnormal value due to an observation error or the like. It is preferable that the diagnostic information observation value is stored.

  The distance calculation unit 140 uses the CPU 911 to calculate an index distance based on the plurality of diagnostic information normal values stored in the normal value storage unit 130 and the diagnostic information observation values stored in the observation value storage unit 120. The index distance is a numerical value representing how far the diagnostic information observation value stored in the observation value storage unit 120 is away from the set formed by a plurality of normal values of diagnostic information stored in the normal value storage unit 130. The larger the index distance, the more the diagnosis target device 810 is more deteriorated. Details of the index distance will be described later.

  The deterioration degree determination unit 150 uses the CPU 911 to determine the deterioration degree of the diagnosis target device 810 based on the index distance calculated by the distance calculation unit 140. The deterioration degree determination unit 150 may be configured to divide the deterioration degree into a relatively small number of stages, such as “normal”, “light deterioration”, and “heavy deterioration”. Alternatively, the deterioration degree determination unit 150 may be configured to represent the deterioration degree with a numerical value of 0 to 100, for example.

The deterioration degree storage unit 160 stores the degree of deterioration determined by the deterioration degree determination unit 150 using the CPU 911.
The deterioration degree storage unit 160 may be configured to store only the deterioration degree determined last by the deterioration degree determination unit 150. Alternatively, the deterioration degree storage unit 160 may be configured to store a plurality of deterioration degrees determined by the deterioration degree determination unit 150 within a predetermined period. Alternatively, the size of the storage area for storing the deterioration degree may be determined in advance, and the deterioration degree storage unit 160 may be configured to store as many deterioration degrees as possible in the storage area as a history.

  Using the CPU 911, the deterioration degree notification unit 170 notifies the administrator of the diagnosis target device 810 of the deterioration degree determined by the deterioration degree determination unit 150. For example, the deterioration degree notification unit 170 displays a message indicating the deterioration degree determined by the deterioration degree determination unit 150 using the display device 901. Alternatively, the deterioration degree notification unit 170 uses the CPU 911 to generate an e-mail that notifies the deterioration degree and transmits it to the mail address of the administrator of the diagnosis target device 810. Note that the degradation level notification unit 170 may be configured to change the notification method according to the degradation level determined by the degradation level determination unit 150. For example, when the deterioration degree determined by the deterioration degree determination unit 150 is “normal”, the deterioration degree notification unit 170 may not be configured to notify the deterioration degree.

When the observation device 820 observes the diagnosis target device 810 based on an instruction from the observation value acquisition unit 110 and outputs the diagnosis information observation value, the observation value acquisition unit 110 stores the deterioration stored in the deterioration degree storage unit 160. The configuration may be such that the observation period and the observation item are changed based on the degree.
For example, the observed value acquisition unit 110 uses the CPU 911 to acquire the latest deterioration level stored in the deterioration level storage unit 160. If the acquired deterioration degree is "normal", the observation value acquisition unit 110, using the CPU 911, the observation period and _{T 1,} the observation items and two _I 1, _{I 2.} When the acquired deterioration degree is “light deterioration”, the observation value acquisition unit 110 uses the CPU 911 to set the observation period to T ₂ shorter than T ₁ and the observation items I ₁ , I ₂ , I ₃ , I _4. 4 types. When the acquired deterioration degree is “heavy deterioration”, the observation value acquisition unit 110 uses the CPU 911 to set the observation cycle to T _{3 which} is shorter than T ₂ and further increase the observation items to 10 types.

  Next, the index distance will be described.

ただし、Ｘは、各観測項目についての診断情報観測値を成分とする列ベクトルである。μは、各観測項目についての診断情報正常値の平均値を成分とする列ベクトルである。右肩のＴは、転置を表わす。Σは、各観測項目についての診断情報正常値の共分散行列である。右肩の−１は、逆行列を表わす。 The index distance calculated by the distance calculation unit 140 is, for example, the Mahalanobis distance. The Mahalanobis distance is defined by the following equation:

Here, X is a column vector whose component is a diagnostic information observation value for each observation item. μ is a column vector whose component is an average value of normal values of diagnostic information for each observation item. T on the right shoulder represents transposition. Σ is a covariance matrix of normal values of diagnostic information for each observation item. -1 on the right shoulder represents an inverse matrix.

For example, the normal value storage unit 130 stores diagnostic information normal values for a plurality of observation items, for example. Let m be the number of observation items of diagnostic information normal values stored in the normal value storage unit 130. m is an integer of 1 or more. The normal value storage unit 130 stores diagnostic information normal values observed simultaneously for m observation items as one set. That is, the normal value storage unit 130 stores a set of m diagnostic information normal values. Let n be the number of sets of diagnostic information normal values stored in the normal value storage unit 130. n is an integer of 1 or more. That is, the normal value storage unit 130 stores m × n diagnostic information normal values in total.
The distance calculation unit 140 calculates a vector μ and an inverse matrix Σ ⁻¹ of the covariance matrix Σ based on m × n diagnostic information normal values stored in the normal value storage unit 130 in advance.
For example, the distance calculation unit 140 uses the CPU 911 to acquire n diagnostic information normal values stored in the normal value storage unit 130 for each of m observation items. The distance calculation unit 140 uses the CPU 911 to calculate a total value of the acquired n diagnostic information normal values. Using the CPU 911, the distance calculation unit 140 divides the calculated total value by n to calculate the average value of the normal diagnostic information value. The distance calculation unit 140 uses m average values calculated for m observation items as components of the m-th order vector μ.
Further, the distance calculation unit 140 uses the CPU 911 to obtain m diagnostic information normal values stored in the normal value storage unit 130 for each of n sets of normal values of diagnostic information stored in the normal value storage unit 130. To do. The distance calculation unit 140 uses the CPU 911 to subtract the average value of the diagnostic information normal value from the diagnostic information normal value for each of the acquired m diagnostic information normal values, and to obtain a deviation from the average diagnostic information normal value. Is calculated. The distance calculation unit 140 uses the CPU 911 to calculate the product of the two deviations calculated for the two observation items for each combination that selects the two observation items from the m observation items. However, the two observation items may be the same observation item. Therefore, there are (m × (m + 1) / 2) combinations for selecting two observation items from m observation items. That is, the distance calculating unit 140 calculates (n × m × (m + 1) / 2) products in total. Using the CPU 911, the distance calculation unit 140 calculates the total value of the calculated n products for each of (m × (m + 1) / 2) combinations. The distance calculation unit 140 uses the CPU 911 to calculate the covariance by dividing the calculated total value by n. The distance calculation unit 140 uses (m × (m + 1) / 2) covariances calculated for (m × (m + 1) / 2) combinations as components of the m-th order covariance matrix Σ. The distance calculation unit 140 uses the CPU 911 to calculate an inverse matrix Σ ⁻¹ of the m-th order covariance matrix Σ having (m × (m + 1) / 2) covariances as components.

The distance calculation unit 140 calculates the Mahalanobis distance every time the observation value acquisition unit 110 acquires the diagnostic information observation value. For example, the distance calculation unit 140 uses the CPU 911 to acquire m diagnostic information observation values simultaneously observed by the observation device 820 for m observation items from the observation value storage unit 120. The distance calculation unit 140 uses the CPU 911 to subtract the average value of the diagnostic information normal value calculated in advance for the observation item from the acquired diagnostic information observation value for each of the m observation items, to obtain the diagnostic information observation value. The deviation from the average value of the diagnostic information normal value is calculated. The distance calculation unit 140 uses the CPU 911 to multiply the m-th column vector having m deviations calculated for m observation items as components from the left by the m-th order inverse matrix Σ ⁻¹ calculated in advance from the left. , M-th column vector is calculated. The distance calculation unit 140 uses the CPU 911 to multiply the m-th row vector having m deviations calculated for m observation items as a component by the calculated m-th column vector from the right, Calculate a scalar value. The distance calculation unit 140 uses the CPU 911 to calculate the parallel root of the calculated scalar value and sets it as the Mahalanobis distance.

  The distance calculation unit 140 may be configured to calculate not the Mahalanobis distance but the square of the Mahalanobis distance as the index distance. In that case, it is not necessary to perform the process of calculating the last parallel root among the processes described above, so that the amount of calculation can be reduced.

  In addition, the distance calculation unit 140 is not configured to calculate the index distance based on the latest diagnostic information observation value acquired by the observation value acquisition unit 110, but is a plurality of times of diagnosis information observation recently acquired by the observation value acquisition unit 110. The index distance may be calculated based on the value.

For example, the distance calculation unit 140 uses the CPU 911 to acquire diagnostic information observation values for a predetermined number of times in order from the latest observation time among the diagnosis information observation values stored in the observation value storage unit 120.
Alternatively, the distance calculation unit 140 uses the CPU 911 to set the observation time as the reference time from among the diagnostic information observation values stored in the observation value storage unit 120 based on the time that has been traced back by a predetermined time from the current time. Diagnostic information observation values obtained later are acquired.

For example, the distance calculation unit 140 uses the CPU 911 to calculate an average value for each observation item based on the acquired diagnosis information observation values for a plurality of times. The distance calculation unit 140 uses the CPU 911 to calculate an index distance based on the calculated average value for each observation item.
Alternatively, the distance calculation unit 140 uses the CPU 911 to calculate an index distance for each acquired diagnostic information observation value. The distance calculation unit 140 uses the CPU 911 to calculate an average value of the plurality of calculated index distances and set it as the index distance as a whole.

The distance calculation unit 140 may be a method of changing the index distance calculation method based on the deterioration degree stored in the deterioration degree storage unit 160.
For example, the distance calculation unit 140 uses the CPU 911 to acquire the latest deterioration level stored in the deterioration level storage unit 160. When the acquired degree of deterioration is “normal”, the distance calculation unit 140 uses the CPU 911 to calculate the index distance using a predetermined first calculation method. When the acquired degree of deterioration is “light deterioration” or “heavy deterioration”, the distance calculation unit 140 uses the CPU 911 to calculate the index distance using a predetermined second calculation method. Here, the second calculation method is a calculation method in which the degree of deterioration is higher than that of the first calculation method, but the calculation amount for calculating the index distance is increased accordingly. For this reason, when the deterioration of the diagnosis target device 810 has not progressed, the amount of calculation required for calculating the index distance can be suppressed, and when the deterioration of the diagnosis target device 810 has progressed to some extent, the degree of deterioration can be increased. It can be determined accurately.

  When the deterioration degree determination unit 150 is configured to determine the deterioration degree in a relatively small number of stages such as “normal”, “light deterioration”, and “heavy deterioration”, the deterioration degree determination unit 150 uses the CPU 911, for example, The index distance calculated by the distance calculation unit 140 is compared with a predetermined threshold of 1 or more. For example, when the index distance is smaller than the first threshold, the deterioration degree determination unit 150 determines “normal”, and when the index distance is between the first threshold and the second threshold larger than that, the degree of deterioration is determined. The determination unit 150 determines “light deterioration”, and when the index distance is larger than the second threshold, the deterioration degree determination unit 150 determines “heavy deterioration”.

When the distance calculation unit 140 is configured to calculate the index distance as a whole by calculating a plurality of index distances and taking the average, the distance calculation unit 140 further uses the CPU 911 to further calculate the index distance. A configuration for calculating variance or standard deviation may be used. The deterioration degree determination unit 150 may be configured to calculate the reliability of the deterioration degree determination based on the variance or standard deviation calculated by the distance calculation unit 140 using the CPU 911.
For example, assuming that the index distance is a sample value collected from a population according to a normal distribution using the CPU 911, the deterioration degree determination unit 150 has a probability that the average value of the population is smaller than the first threshold value, A probability that the average value of the population is larger than the first threshold value and smaller than the second threshold value and a probability that the average value of the population value is larger than the second threshold value are calculated. Using the CPU 911, the deterioration degree determination unit 150 compares the calculated three probabilities and obtains the highest probability. When the probability that the average value of the population is smaller than the first threshold is the highest, the deterioration degree determination unit 150 uses the CPU 911 to determine “normal”. When the probability that the average value of the population is larger than the first threshold and smaller than the second threshold is the largest, the deterioration degree determination unit 150 uses the CPU 911 to determine “light deterioration”. When the probability that the average value of the population is greater than the second threshold is the highest, the deterioration degree determination unit 150 uses the CPU 911 to determine “heavy deterioration”. Further, the deterioration degree determination unit 150 uses the CPU 911 to set the calculated probability as the reliability of the determination.

  The deterioration degree storage unit 160 stores the determination reliability calculated by the deterioration degree determination unit 150 together with the deterioration degree determined by the deterioration degree determination unit 150 using the magnetic disk device 920.

  Using the CPU 911, the deterioration degree notification unit 170 notifies the reliability of the determination calculated by the deterioration degree determination unit 150 together with the deterioration degree determined by the deterioration degree determination unit 150. Note that the deterioration degree notification unit 170 is configured to change the notification method of the deterioration degree and the reliability, or to change whether or not to notify, depending on the determination reliability calculated by the deterioration degree determination unit 150. Also good. For example, when the reliability of the determination calculated by the deterioration level determination unit 150 is lower than a predetermined threshold, the deterioration level and the reliability level are not notified so as not to confuse the administrator of the diagnosis target device 810 with low reliability information. There may be. Conversely, when the reliability is low, there is a possibility that the observation device 820 has an abnormality. Therefore, even if the determination by the deterioration degree determination unit 150 is “normal”, the deterioration degree notification unit 170 uses the CPU 911, The configuration may be such that a notification is made by the same notification method as in the case of “heavy deterioration”.

The observation value acquisition unit 110 may be configured to change the observation period and the observation item according to the determination reliability calculated by the deterioration degree determination unit 150. For example, when the reliability of determination is lower than a predetermined threshold, the observation value acquisition unit 110 uses the CPU 911 to shorten the observation period and increase the number of acquisitions of diagnostic information observation values compared to when the reliability is high. Or when the reliability of determination is lower than a predetermined threshold value, the observation value acquisition unit 110 uses the CPU 911 to increase the number of observation items as compared with the case where the reliability is high.
Similarly, the distance calculation unit 140 may be configured to change the index distance calculation method according to the determination reliability calculated by the deterioration determination unit 150. For example, when the determination reliability is lower than a predetermined threshold, the distance calculation unit 140 changes the calculation method of the index distance so that the calculation amount is larger than that when the reliability is high, but the accuracy of the deterioration determination is high. To.

FIG. 9 is a flowchart showing an example of the flow of the deterioration degree determination process S510 in this embodiment.
In the deterioration level determination process S510, the deterioration level determination apparatus 100 determines the deterioration level of the diagnosis target apparatus 810. The deterioration degree determination process S510 includes an observation value acquisition step S511, an index distance calculation step S512, an index distance average step S513, a deterioration degree calculation step S514, and a deterioration degree notification step S515.

In the observation value acquisition step S511, the observation value acquisition unit 110 calculates an observation period and an observation item using the CPU 911 based on the deterioration degree and reliability stored in the deterioration degree storage unit 160. The observation value acquisition unit 110 calculates the next observation time using the CPU 911 based on the time when the diagnosis information observation value was acquired last time and the calculated observation period. Using the CPU 911, the observation value acquisition unit 110 stands by until the calculated observation time is reached.
When the observation time is reached, the observation value acquisition unit 110 uses the CPU 911 to instruct the observation device 820 to observe the calculated observation item.
The diagnosis target device 810 observes the diagnosis target device 810 according to an instruction from the observation value acquisition unit 110. The diagnosis target device 810 outputs the observed diagnostic information observation value.
The observation value acquisition unit 110 uses the CPU 911 to acquire the diagnostic information observation value output from the diagnosis target device 810. The observation value storage unit 120 stores the diagnostic information observation values acquired by the observation value acquisition unit 110 using the magnetic disk device 920.

In the index distance calculation step S512, the distance calculation unit 140 determines the index distance calculation method using the CPU 911 based on the deterioration degree and reliability stored in the deterioration degree storage unit 160. The distance calculation unit 140 uses the CPU 911 to determine whether a value that can be calculated in advance, such as an average value of normal values of diagnosis information, has been calculated in the determined calculation method. If it is determined that the calculation has not been completed, the distance calculation unit 140 uses the CPU 911 to calculate those values. The distance calculation unit 140 stores the calculated value using the magnetic disk device 920.
The distance calculation unit 140 calculates the index distance by the determined calculation method using the CPU 911 based on the diagnosis information observation value stored in the observation value storage unit 120 in the observation value acquisition step S511. The distance calculation unit 140 stores the calculated index distance using the magnetic disk device 920.

  In the index distance averaging step S513, the distance calculation unit 140 uses the CPU 911 to observe the index distance for the past predetermined number of times or the past predetermined period from the index distances stored in the index distance calculation step S512. The index distance calculated based on the diagnostic information observation value is acquired. The distance calculation unit 140 uses the CPU 911 to calculate the average value and variance of the acquired index distances.

  In the deterioration degree calculation step S514, the deterioration degree determination unit 150 uses the CPU 911 to determine the deterioration degree of the diagnosis target device 810 based on the average value and variance of the index distances calculated by the distance calculation unit 140 in the index distance average step S513. And the reliability is calculated. The deterioration degree storage unit 160 stores the deterioration degree and reliability determined by the deterioration degree determination unit 150 using the magnetic disk device 920.

In the deterioration degree notification step S515, the deterioration degree notification unit 170 determines a notification method using the CPU 911 based on the deterioration degree and the reliability stored in the deterioration degree storage unit 160 in the deterioration degree calculation step S514. The deterioration degree notification unit 170 notifies the deterioration degree and reliability stored in the deterioration degree storage unit 160 in the deterioration degree calculation step S514 using the CPU 911 according to the determined notification method.
The degradation degree determination apparatus 100 returns the process to the observation value acquisition step S511, and acquires the next diagnosis information observation value.

Embodiment 5 FIG.
The fifth embodiment will be described with reference to FIGS.
In addition, about the part which is common in Embodiment 4, the same code | symbol is attached | subjected and description is abbreviate | omitted.

FIG. 10 is a system configuration diagram showing an example of the overall configuration of the deterioration degree determination system 800 in this embodiment.
The degradation degree determination system 800 includes a plurality of diagnosis target devices 810a to 810z. The plurality of diagnosis target devices 810a to 810z are devices that are considered to have the same relationship between the diagnostic information observation value and the degree of deterioration. For example, the diagnosis target devices 810a to 810z are disk drive devices of the same manufacturer and model.
The diagnosis target devices 810a to 810z can always be increased or decreased. For example, a deteriorated diagnosis target device may be removed from the system, or a new diagnosis target device may be added to the system.

  The observation device 820 observes the plurality of diagnosis target devices 810a to 810z and outputs diagnostic information observation values for the respective diagnosis target devices 810a to 810z. The plurality of observation devices 820 may be configured to observe the corresponding diagnosis target devices.

In addition to the configuration described in the fourth embodiment, degradation level determination apparatus 100 further includes a promotion determination unit 180.
The promotion determination unit 180 uses the CPU 911 to determine whether the diagnosis target devices 810a to 810z have deteriorated when the observation device 820 observes the diagnostic information observation values stored in the observation value storage unit 120. The determination method will be described later. When the promotion determination unit 180 determines that the diagnosis information observation value is observed when the diagnosis target devices 810a to 810z are not deteriorated, the diagnosis information observation value is promoted to a diagnosis information normal value.
The normal value storage unit 130 uses the magnetic disk device 920 to store the diagnostic information observation value determined by the promotion determination unit 180 as the diagnostic information observation value observed when the diagnosis target devices 810a to 810z are not deteriorated. Store as normal value of diagnostic information.

  A determination method of the promotion determination unit 180 will be described.

  For example, if the diagnosis target apparatus operates normally for a longer time than a predetermined deterioration determination time (for example, 1000 hours), it is considered that the diagnosis target apparatus has not deteriorated at least before that time.

For example, identification information unique to each of the diagnosis target devices 810a to 810z, such as a serial number, is one of the diagnostic information observation values. The observation device 820 observes identification information of the diagnosis target devices 810a to 810z as diagnosis information observation values. The observation value acquisition unit 110 acquires the identification information observed by the observation device 820 as the diagnostic information observation value. The observation value storage unit 120 stores the identification information acquired by the observation value acquisition unit 110 as the diagnostic information observation value.
The promotion determination unit 180 determines a diagnosis target device that is currently operating based on the identification information stored in the observation value storage unit 120. If there is a diagnostic information observation value observed before the reference time for a diagnostic target device with the same identification information, using the time before the deterioration determination time as the reference time, the diagnostic target at that point The apparatus determines that it has not deteriorated.
The promotion determination unit 180 uses the CPU 911 to acquire identification information of the currently operated diagnostic target device from the diagnostic information observation value stored in the observation value storage unit 120. The promotion determination unit 180 has the same identification information as the acquired identification information among the diagnosis information observation values stored in the observation value storage unit 120, and the diagnosis information whose elapsed time from the observation time is greater than a predetermined time. Get observations. The normal value storage unit 130 stores the diagnostic information observation value acquired by the promotion determination unit 180 as the diagnostic information normal value.

  Whether or not the diagnostic target device is currently operating is not determined based on the identification information stored in the observation value storage unit 120. For example, an administrator of the diagnostic target devices 810a to 810z replaces the diagnostic target device. In this case, the determination may be made by inputting to the deterioration degree determination apparatus 100. For example, when the administrator changes the configuration of the diagnosis target devices 810a to 810z, the promotion determination unit 180 uses the CPU 911 to acquire information input by the administrator. The promotion determination unit 180 uses the magnetic disk device 920 to store the acquired information. The promotion determination unit 180 uses the CPU 911 to determine the operation start time and the operation stop time for each diagnosis target device 810a to 810z based on the stored information.

  Further, the promotion determination unit 180 may be configured to change the determination method based on the deterioration level and the reliability stored in the deterioration level storage unit 160. For example, when the latest deterioration degree stored in the deterioration degree storage unit 160 for the diagnosis target device is “light deterioration”, the promotion determination unit 180 uses the CPU 911 to determine the deterioration determination time more than when it is “normal”. Lengthen. Alternatively, when the latest deterioration degree stored in the deterioration degree storage unit 160 for the diagnosis target device is “heavy deterioration”, the promotion determination unit 180 uses the CPU 911 to obtain the diagnosis information observation value for the diagnosis target device. Don't promote.

  The normal value storage unit 130 stores the diagnostic information normal value in advance using the magnetic disk device 920, and additionally stores the diagnostic information normal value according to the determination result of the promotion determination unit 180.

  Alternatively, the normal value storage unit 130 may be configured not to store diagnostic information normal values at the start of operation of the deterioration degree determination system 800. In this case, if the number of normal values of diagnostic information stored in the normal value storage unit 130 is small, the accuracy of determination by the deterioration level determination unit 150 is lowered. Therefore, when the number of normal values of diagnostic information is small, temporary operation is performed. For example, the deterioration degree notification unit 170 uses the CPU 911 to compare the number of diagnostic information normal values stored in the normal value storage unit 130 with a predetermined threshold. When the number of diagnostic information normal values is less than the threshold value, the deterioration degree notification unit 170 does not notify the deterioration degree using the CPU 911. Alternatively, when the number of normal values of diagnosis information is less than the threshold value, the deterioration degree notification unit 170 notifies that the deterioration degree determination system 800 is in temporary operation using the CPU 911.

FIG. 11 is a flowchart showing an example of the flow of the promotion process S520 in this embodiment.
In the promotion process S520, the degradation degree determination apparatus 100 promotes the diagnostic information observation value to the diagnostic information normal value. The promotion process S520 includes an identification acquisition step S521, a determination time calculation step S522, an observation time acquisition step S523, a promotion determination step S524, an erasure determination step S525, a promotion step S526, and an erasure step S527.
The degradation degree determination apparatus 100 repeatedly executes the promotion process S520 at a predetermined cycle. The execution period of the promotion process S520 is longer than the observation period in which the observation apparatus 820 observes the diagnosis target apparatuses 810a to 810z, and is, for example, about 5 to 10 times the observation period.

In the identification acquisition step S521, the promotion determination unit 180 uses the CPU 911, and the observation time is later than the time when the promotion processing S520 was previously executed from the set of diagnostic information observation values stored in the observation value storage unit 120. One set of diagnostic information observation values is selected.
When all the sets of diagnostic information observation values after the previous execution time of the promotion process S520 have been selected and there is no set of diagnostic information observation values to be selected, the promotion determination unit 180 uses the CPU 911 to The process proceeds to observation time acquisition step S523.
If there is a set of unselected diagnostic information observation values among the diagnostic information observation values whose observation time is earlier than the previous execution time of the promotion process S520, the promotion determination unit 180 uses the CPU 911 to perform unselected diagnostic information observation. One set of diagnostic information observation values is selected from the set of values. The promotion determination unit 180 uses the CPU 911 to acquire identification information from the selected set of diagnostic information observation values. Using the CPU 911, the promotion determination unit 180 proceeds to the determination time calculation step S522.

In the determination time calculation step S522, the promotion determination unit 180 uses the CPU 911 to determine whether or not the deterioration determination time has been calculated for the diagnosis target device identified by the identification information acquired in the identification acquisition step S521. . If it is determined that the degradation determination time has not yet been calculated, the promotion determination unit 180 uses the CPU 911 to determine whether the degradation level determination unit 150 for the diagnosis target device uses the degradation level stored in the degradation level storage unit 160. Acquire the latest degree of degradation determined. Using the CPU 911, the promotion determination unit 180 calculates a deterioration determination time based on the acquired deterioration degree. Using the magnetic disk device 920, the promotion determination unit 180 stores a set of identification information and the calculated deterioration determination time.
Using the CPU 911, the promotion determination unit 180 returns the process to the identification acquisition step S521, and selects the next set of diagnostic information observation values.

In the observation time acquisition step S523, the promotion determination unit 180 uses the CPU 911 to select one set of diagnostic information observation values from the set of diagnostic information observation values stored in the observation value storage unit 120.
When all the sets of diagnostic information observation values stored in the observation value storage unit 120 have been selected and there is no set of diagnostic information observation values to be selected, the promotion determination unit 180 uses the CPU 911 to execute the promotion process S520. finish.
When there is a set of unselected diagnostic information observation values among the set of diagnostic information observation values stored in the observation value storage unit 120, the promotion determination unit 180 uses the CPU 911 to store the unselected diagnostic information observation values. One set of diagnostic information observation values is selected from the set. Using the CPU 911, the promotion determination unit 180 advances the process to the promotion determination step S524.

In the promotion determination step S524, the promotion determination unit 180 uses the CPU 911 to acquire identification information from the set of diagnostic information observation values selected in the observation time acquisition step S523. Using the CPU 911, the promotion determination unit 180 acquires a set in which the identification information is the same as the acquired identification information from the combination of the identification information stored in the determination time calculation step S522 and the deterioration determination time.
If there is no combination in which the identification information is the same as the acquired identification information in the combination of the identification information stored in the determination time calculation step S522 and the deterioration determination time, the promotion determination unit 180 uses the CPU 911 to delete The process proceeds to determination step S525.
When there is a pair in which the identification information is the same as the acquired identification information in the pair of the identification information stored in the determination time calculation step S522 and the deterioration determination time, the promotion determination unit 180 uses the CPU 911 to perform observation. The elapsed time from the observation time is longer than the deterioration determination time based on the observation time observed by the observation device 820 for the set of diagnostic information observation values selected in the time acquisition step S523, the acquired deterioration determination time, and the current time. Or whether it is short.
When the elapsed time from the observation time is longer than the deterioration determination time, the promotion determination unit 180 uses the CPU 911 to advance the process to the promotion step S526.
When the elapsed time from the observation time is shorter than the deterioration determination time, the promotion determination unit 180 uses the CPU 911 to return the process to the observation time acquisition step S523 and selects the next set of diagnostic information observation values.

In the erasure determination step S525, the promotion determination unit 180 uses the CPU 911 to perform observation based on the observation time observed by the observation device 820 for the set of diagnostic information observation values selected in the observation time acquisition step S523 and the current time. It is determined whether the elapsed time from the time is longer or shorter than a predetermined erasure determination time. The erasure determination time is a time longer than the maximum value of the deterioration determination time calculated in the determination time calculation step S522.
When the elapsed time from the observation time is longer than the erasure determination time, the promotion determination unit 180 advances the process to the erasure process S527 using the CPU 911.
When the elapsed time from the observation time is shorter than the erasure determination time, the promotion determination unit 180 uses the CPU 911 to return the process to the observation time acquisition step S523 and selects the next set of diagnostic information observation values.

  In the promotion step S526, the normal value storage unit 130 uses the magnetic disk device 920 to store the set of diagnostic information observation values selected by the promotion determination unit 180 in the observation time acquisition step S523 as a set of diagnostic information normal values. .

In the erasing step S527, the observation value storage unit 120 selects the set of diagnostic information observation values selected by the promotion determination unit 180 in the observation time acquisition step S523 from the set of fleet information observation values stored using the magnetic disk device 920. to erase.
Using the CPU 911, the promotion determination unit 180 returns the process to the observation time acquisition step S523 and selects the next set of diagnostic information observation values.

  As described above, the configuration described in each embodiment is an example, and a configuration obtained by partially modifying the configuration may be used. For example, the structure which combined the structure demonstrated in different embodiment may be sufficient. Or it is good also as a structure which replaced the structure of the part which is not main with the other structure.

The deterioration level determination device (100; disk drive deterioration level determination device 2) described above includes a processing device (CPU 911) that processes data, a storage device (such as a magnetic disk device 920) that stores data, and an observation value acquisition unit. (110; diagnosis information collection unit 5), normal value storage unit (130; normal sample storage unit 6), distance calculation unit (140; diagnosis information distance calculation unit 7), and deterioration degree determination unit (150; disk drive) Deterioration degree determination unit 8).
The observation value acquisition unit (110; 5) uses the processing device (911) to acquire a numerical value that changes as the diagnosis target device (810; disk drive 4) deteriorates as a diagnostic information observation value.
The normal value storage unit (130; 6) uses the storage device (920) to store a plurality of diagnostic information observation values as diagnostic information normal values when the diagnostic target device (810; 4) is not deteriorated. .
The distance calculation unit (140; 7) converts the diagnosis information observation value acquired by the observation value acquisition unit (110; 5) and the plurality of diagnosis information normal values stored by the normal value storage unit (130; 6). Based on this, the processing device (911) is used to calculate, as an index distance, a numerical value representing how far the diagnostic information observation value is from a set formed by a plurality of normal values of the diagnostic information.
The deterioration degree determination unit (150; disk drive deterioration degree determination unit 8) uses the processing device (911) based on the index distance calculated by the distance calculation unit (140; 7), and the diagnosis target device ( 810; The deterioration degree of 4) is determined.

  Thereby, even if it is not known in advance how the diagnostic information observation value changes when the diagnostic target device progresses, the degree of degradation of the diagnostic target device can be determined based on the diagnostic information observation value. it can.

The observation value acquisition unit (110; 5) uses the processing device (911) to acquire a plurality of diagnostic information observation values with different observation times.
The distance calculation unit (140; 7) uses the processing device (911) to determine how far a set formed by the plurality of diagnostic information observation values is away from a set formed by the plurality of normal values of the diagnosis information. The index distance to represent is calculated.

  Thereby, even when an abnormal value due to an observation error or the like is included in the diagnosis information observation value, it is possible to determine the deterioration of the diagnosis target device.

  The observation value acquisition unit (110; 5) uses the processing device (911) to set an acquisition period for acquiring the diagnostic information observation value based on the degree of deterioration determined by the deterioration degree determination unit (150; 8). The acquisition period is shortened as the deterioration degree increases.

  Since the acquisition cycle is shortened as the deterioration of the observation target device proceeds, it is possible to quickly determine the progress of the deterioration.

  The distance calculation unit (140; 7) uses the processing device (911) to calculate a test value by a statistical method as the index distance.

  This makes it possible to make a highly reliable determination backed by statistical theory.

  The observation value acquisition unit (110; 5) acquires a plurality of types of diagnostic information observation values using the processing device (911).

  By determining the degree of deterioration using a plurality of types of diagnostic information observation values, the determination accuracy can be increased.

  The observation value acquisition unit (110; 5) uses the processing device (911) to determine the acquisition type of the diagnostic information observation value to be acquired based on the deterioration level determined by the deterioration level determination unit (150; 8). The acquisition type is increased as the degree of deterioration increases.

  Since the number of acquisition types increases as the deterioration of the observation target device progresses, it is possible to accurately determine the progress of the deterioration.

  The deterioration degree determination unit (150; 8) uses the processing device (911) to determine that the degree of deterioration of the diagnosis target device is normal when the index distance is smaller than a first threshold, and the index When the distance is larger than the first threshold and smaller than the second threshold, it is determined that the degree of deterioration of the diagnosis target device is slight deterioration. When the index distance is larger than the second threshold, the diagnosis target It is determined that the degree of deterioration of the apparatus is severe deterioration.

  By dividing the degree of deterioration into three, the progress of the deterioration of the diagnosis target apparatus can be easily communicated to the user.

  The deterioration degree determination unit (150; 8) uses the processing device (911) to change the second threshold based on the seriousness determination ratio determined to be severe deterioration by a plurality of determinations, As the severity determination ratio increases, the second threshold value is increased.

  If the second threshold value is increased, the ratio of diagnosis target devices determined to be severely deteriorated decreases, so that the ratio of diagnosis target devices determined to be severely deteriorated can be set to an appropriate ratio. .

The diagnosis target device (810; 4) is, for example, a disk drive device.
The observation value acquisition unit (110; 5) uses the processing device (911) to acquire the SMART value output from the disk drive device as the diagnostic information observation value.

  Since the function of the existing disk drive device is used, the diagnostic information observation value can be easily obtained.

  The deterioration degree determination apparatus (100; 2) described above can be realized by executing a computer program that causes a computer to function as the deterioration degree determination apparatus.

  DESCRIPTION OF SYMBOLS 1 Computer, 2 Disk drive deterioration degree determination apparatus, 3 Operating system, 4 Disk drive, 5 Diagnostic information collection part, 6 Normal sample value storage part, 7 Diagnostic information distance calculation part, 8 Disk drive deterioration degree determination part, 9 Disk drive Deterioration degree notification unit, 10 Diagnostic information collection period calculation unit, 11 Collected diagnosis information change unit, 12 Deterioration degree notification method change unit, 13 Diagnostic information and deterioration degree determination result storage unit, 14 Diagnostic information distance calculation method and deterioration degree determination method Dynamic change unit, 100 degradation level determination device, 110 observation value acquisition unit, 120 observation value storage unit, 130 normal value storage unit, 140 distance calculation unit, 150 degradation level determination unit, 160 degradation level storage unit, 170 degradation level notification , 180 promotion determination unit, 800 degradation degree determination system, 810 diagnosis target device, 820 observation device Device, 901 display device, 902 keyboard, 903 mouse, 904 FDD, 905 CDD, 906 printer device, 907 scanner device, 910 system unit, 911 CPU, 912 bus, 913 ROM, 914 RAM, 915 communication device, 920 magnetic disk device , 921 OS, 922 window system, 923 program group, 924 file group, 931 telephone, 932 facsimile machine, 940 Internet, 941 gateway, 942 LAN.

Claims (11)

A processing device that processes data, a storage device that stores data, an observation value acquisition unit, a normal value storage unit, a distance calculation unit, and a deterioration degree determination unit;
The observation value acquisition unit acquires, as the diagnostic information observation value, a numerical value that changes as the diagnosis target device deteriorates, using the processing device.
The normal value storage unit uses the storage device to store a plurality of diagnostic information observation values as diagnostic information normal values in a state where the diagnosis target device is not deteriorated,
The distance calculation unit uses the processing device to perform a plurality of diagnosis based on the diagnosis information observation value acquired by the observation value acquisition unit and the plurality of diagnosis information normal values stored by the normal value storage unit. A numerical value indicating how far the diagnostic information observation value is from the set of normal information values is calculated as an index distance,
The degradation level determination unit, wherein the degradation level determination unit determines the degradation level of the diagnosis target device using the processing device based on the index distance calculated by the distance calculation unit. The observation value acquisition unit acquires a plurality of diagnostic information observation values with different observation times using the processing device,
The distance calculation unit calculates, using the processing device, an index distance that indicates how far a set formed by the plurality of diagnostic information observation values is apart from a set formed by the plurality of normal values of the diagnostic information. The degradation degree determination apparatus according to claim 1, wherein   The observation value acquisition unit changes the acquisition cycle for acquiring diagnostic information observation values based on the degree of deterioration determined by the deterioration degree determination unit using the processing device, and the acquisition period increases as the degree of deterioration increases. The deterioration degree determination apparatus according to claim 1, wherein the deterioration degree determination apparatus according to claim 1 is shortened.   The degradation degree determination apparatus according to any one of claims 1 to 3, wherein the distance calculation unit calculates a test value by a statistical method as the index distance using the processing device.   The deterioration level determination device according to any one of claims 1 to 4, wherein the observation value acquisition unit acquires a plurality of types of diagnostic information observation values using the processing device.   The observation value acquisition unit changes the acquisition type of the diagnostic information observation value to be acquired based on the deterioration degree determined by the deterioration degree determination unit using the processing device, and the acquisition type increases as the deterioration degree increases. The deterioration degree determination apparatus according to claim 5, wherein the degree of deterioration is increased.   The deterioration degree determination unit determines that the deterioration degree of the diagnosis target device is normal when the index distance is smaller than a first threshold using the processing device, and the index distance is the first threshold. If it is larger and smaller than the second threshold, it is determined that the degree of deterioration of the diagnosis target device is mild deterioration. If the index distance is larger than the second threshold, the degree of deterioration of the diagnosis target device is severe. The deterioration degree determination apparatus according to claim 1, wherein the deterioration degree determination apparatus is determined to be deterioration.   The degradation level determination unit changes the second threshold based on the severity determination ratio determined to be severe deterioration by multiple determinations using the processing device, and the greater the severity determination ratio, 8. The deterioration degree determination apparatus according to claim 7, wherein the second threshold value is increased. The diagnosis target device is a disk drive device,
The said observation value acquisition part acquires the SMART value which the said disk drive apparatus outputs as a diagnostic information observation value using the said processing apparatus, The Claim 1 thru | or 8 characterized by the above-mentioned. Degradation degree determination device.   When the computer having a processing device that processes data and a storage device that stores data is executed, the computer functions as the deterioration degree determination device according to any one of claims 1 to 9. Computer program. The processing device acquires a numerical value that changes as the diagnostic target device deteriorates as a diagnostic information observation value,
The storage device stores a plurality of diagnostic information observation values as diagnostic information normal values in a state where the diagnostic target device is not deteriorated,
Based on the acquired diagnostic information observation value and the stored plurality of normal diagnostic information values, the processing device determines how far the diagnostic information observation value is from the set of the multiple normal diagnostic information values. Calculate the index value as the index distance,
A deterioration degree determination method, wherein the processing device determines a deterioration degree of the diagnosis target device based on the calculated index distance.

Images