JP2017037626A - Device, method, and program for failure prediction - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a failure prediction technique that allows for predicting failure of a recording medium and preventing loss of data stored in the recording medium.SOLUTION: An HDD failure prediction device 100 has an HDD controller 10 configured to receive commands for writing to or reading from an HDD 300 from a host 200, and to write data to or read data from the HDD 300. An abnormal value DB recording unit 40 stores state information of the recording medium when access target points in the recording medium are accessed in an access pattern specific to the recording medium in increments of a specific unit volume. A controller 30 acquires the state information of the recording medium when the access target points are accessed according to the access pattern in increments of the specific unit volume, and registers the state information of the recording medium in association with the access target points in the abnormal value DB recording unit 40.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a failure prediction technique for a recording medium.

  A hard disk may have a bad sector due to a minute defect on the disk surface, making it impossible to read and write, or repeating a retry operation due to a head failure may significantly reduce the data transfer speed. In addition, a failure may occur and the hard disk itself may fail to start up.

  Patent Document 1 discloses a technique for measuring a hard disk transfer time and detecting a hard disk failure sign from the transfer time. When the transfer time is measured every predetermined time (for example, one week) after product shipment and compared with the transfer time at the time of shipment from the factory, if the difference between the two transfer times satisfies the predetermined condition, Warning notification processing is performed to indicate that there may be a problem with the hard disk in the future.

  Patent Document 2 focuses on the number of occurrences of Realized Sector Count (hereinafter referred to as “alternative sector number”) in SMART (Self-Monitoring, Analysis and Reporting Technology) information of a hard disk drive (HDD), and changes in the number of alternative sectors Describes a technique for determining an abnormality of an HDD from the above.

JP 2011-68109 A JP 2001-6273 A

  In the prior art disclosed in Patent Document 1, a value obtained by multiplying a measurement value of transfer time by a predetermined coefficient is compared with a measurement time at the time of factory shipment, and when the former is long, it is determined as a delay region. A warning is issued when the delay area is greater than or equal to a predetermined threshold. However, no consideration was given to variations in access time depending on the head position of a hard disk drive (HDD). For example, even when accessing the same area (sector) of the HDD, if the rotation waiting time is the maximum, the waiting time for one rotation of the disk is required, and the access time is increased by that time, and the rotation waiting time is reduced. In the case of the minimum, the access time is shortened because the access can be made in the shortest time without waiting for the disk rotation. As described above, even when the same area of the HDD is accessed, the access time until the data read is completed varies depending on the head position before access and the access timing. Since no consideration was given, the access time could not be measured with sufficient accuracy. Therefore, even if failure prediction is performed based on such access time, failure prediction of the HDD cannot be performed with high accuracy.

  In the technique described in Patent Document 2, attention is paid to the number of occurrences of alternative sectors in the SMART information, and failure prediction of a hard disk drive (HDD) is performed. It was not possible to distinguish between the occurrence of alternative sectors due to accidental (accidental) factors and the occurrence of alternative sectors due to more serious factors (such as growth of scratches on the disk surface).

  However, if there are alternative sectors distributed on the disk, if the alternative sector itself is accidentally generated, then if the alternative sector does not increase, the HDD itself returns to normal operation due to the occurrence of the alternative sector. So the obstacles are not growing. On the other hand, for example, when an alternative sector is generated in a continuous sector for one circumference of the disk, there is a high possibility that the head is in contact with the disk and scratches the disk. In such a case, there is a high possibility that the head is in contact with a sector that is close to the area where the alternative sector can be recognized, and the failure is likely to grow over time. It is necessary to take measures to rescue the data immediately after recognizing the possibility of failure. However, in the conventional technology, the location where the alternative sector is generated cannot be specified, and therefore, the alternative sector generated on the HDD disk and the alternative sector generated on the consecutive sectors are troubled by the same number of occurrences with the same weight. There was no choice but to judge the degree of.

  In this way, originally, the failure prediction of the HDD greatly differs depending on the position where the alternative sector has occurred. However, in the prior art, the failure determination is based only on the number of occurrences of the alternative sector. A detailed diagnosis of the abnormality could not be made.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide a failure prediction technique capable of accurately predicting a failure occurrence of a recording medium and preventing loss of data in the recording medium. It is in.

  In order to solve the above-described problem, a failure prediction apparatus according to an aspect of the present invention stores state information of the recording medium when the access target portion is accessed in a specific capacity unit according to a predetermined access pattern for the recording medium. The state information of the recording medium is acquired when the access target part is accessed in the specific capacity unit according to the recording unit and the access pattern, and the state information of the recording medium is associated with the access target part and the state And a control unit registered in the information recording unit.

  Another aspect of the present invention is a failure prediction method. The method includes a step of storing status information of the recording medium when a location to be accessed is accessed in a specific capacity unit according to a predetermined access pattern for the recording medium in a status information recording unit, and the location to be accessed according to the access pattern. Acquiring the status information of the recording medium when accessing the recording medium in the specific capacity unit, and registering the status information of the recording medium in the status information recording unit in association with the access target location.

  It should be noted that any combination of the above-described constituent elements and a conversion of the expression of the present invention between a method, an apparatus, a system, a recording medium, a computer program, etc. are also effective as an aspect of the present invention.

  According to the present invention, it is possible to accurately predict the occurrence of a failure in a recording medium and prevent data loss in the recording medium.

1 is a configuration diagram of an HDD failure prediction apparatus according to a first embodiment. It is a flowchart which shows the failure prediction procedure by the HDD failure prediction apparatus of FIG. It is a flowchart which shows the detailed procedure of the SMRT information determination process of FIG. It is a figure explaining the access pattern with respect to a command issue number. It is a figure which shows the abnormal value database which recorded the command issue number and SMART information correspondingly. It is a figure which shows the error counter counted for every command issue number. 10 is a flowchart illustrating a failure prediction procedure by the HDD failure prediction apparatus according to the second embodiment. It is a flowchart which shows the detailed procedure of the failure prediction process of FIG. It is a flowchart which shows the detailed procedure of the warning determination process of FIG. It is a flowchart which shows the detailed procedure of the error determination process of FIG. It is a figure explaining the access pattern for measuring command execution time. It is a figure explaining the command execution time measured by the access pattern of FIG. It is the graph which showed the schematic diagram of FIG. 12A by the measured value. It is a figure explaining the evaluation data and threshold value data in the segment which grouped the command issue number. FIG. 14A shows an access time when the HDD 300 is normal, and FIG. 14B shows an access time when an abnormality occurs in the HDD 300. FIG. 15A shows an abnormal value database recorded in association with a command issue number and a threshold, and FIG. 15B shows an abnormal value database recorded in association with a segment number and a threshold. is there. FIG. 16A shows a warning counter counted for each command issue number, and FIG. 16B shows an error counter counted for each command issue number. 10 is a flowchart illustrating a detailed procedure of warning determination processing according to the third embodiment. 12 is a flowchart illustrating a detailed procedure of recovery processing according to the third embodiment. It is a figure explaining the storage method of SMART information.

(Embodiment 1)
The occurrence of an alternative sector is a method for recovering the function of the HDD manufacturer as a HDD manufacturer to cut off the sector in which an unrecoverable failure has occurred and switch the HDD to a new alternative sector to return to normal operation. The outbreak itself is not bad. In fact, accidental sector destruction is a phenomenon that can be seen even in the early stage of operation. For such sector destruction, HDDs can be used normally without the user being aware of the occurrence of alternative sectors. You can continue.

  However, for example, in the case of writing in 100 sectors, when occurrence of as many as 10 alternative sectors is confirmed, for example, there is a possibility of physical failure due to contact with the head, and when this is expected, Furthermore, a failure may occur in a nearby area close to the accessed sector, and the failure may grow.

  The SMART information only indicates the number of occurrences as the number of alternative sectors, and the location and timing of occurrence of the alternative sectors are not known. For this reason, simply referring to the SMART information cannot identify the area in which the alternative sector is increasing and for what reason, and cannot predict the failure growth. Therefore, even if an alternative sector has already occurred due to a serious failure due to physical damage, for example, the symptom may progress until data cannot be rescued from the HDD. As described above, the method that pays attention only to the number of occurrences of alternative sectors has a problem in that it is impossible to detect the growth of the HDD failure with high accuracy and to rescue the data before the failure.

  In order to solve this problem, in this embodiment, instead of paying attention to the number of alternative sectors generated after the occurrence of a problem such as an access failure, the alternative sector is always monitored by monitoring the occurrence status of alternative sectors for access. This paper proposes a method for identifying the occurrence position of the disk, making it possible to create the distribution of alternative sectors on the disk, and predicting the failure from the distribution.

  FIG. 1 is a configuration diagram of an HDD failure prediction apparatus 100 according to the embodiment. The HDD failure prediction apparatus 100 has a function of driving the HDD 300 used by the host 200 and a function of predicting a failure of the HDD 300. The HDD failure prediction apparatus 100 includes an HDD controller 10, a temporary storage unit 20, a control unit 30, and an abnormal value DB recording unit 40. These configurations can be realized by hardware, software, or a combination thereof. It should be noted that the HDD failure prediction apparatus 100 and the host 200 can be configured integrally. Further, the HDD failure prediction apparatus 100, the host 200, and the HDD 300 may be configured integrally.

  The HDD controller 10 receives a read / write command for the HDD 300 from the host 200 based on the ATA standard of the hard disk drive, writes data to the HDD 300, and reads data from the HDD 300. The HDD controller 10 issues a SMART information read command. The temporary storage unit 20 temporarily stores the transfer data in a FIFO structure in order to collect the transfer data in a specific capacity unit.

  The control unit 30 extracts the number of alternative sectors from the SMART information read by the HDD controller 10, records it in the abnormal value DB recording unit 40, and the number of alternative sectors so far recorded in the abnormal value DB recording unit 40. The service life of the HDD 300 is determined from the change in.

  The abnormal value DB recording unit 40 registers the number of alternative sectors extracted from the SMART information for each command issue number.

  The host 200 includes a display unit and an input unit (not shown). The input to the input unit of the host 200 is transmitted to the control unit 30 of the HDD failure prediction apparatus 100 and processed. The control unit 30 of the HDD failure prediction apparatus 100 also functions as a display control unit that controls the display unit of the host 200.

  When the host 200 issues a data write command to the HDD 300, the written data is temporarily stored in the temporary storage unit 20 via the HDD controller 10. When the temporarily stored write data reaches a specific capacity, the HDD controller 10 writes the old data on the time axis from the old data to the HDD 300 according to the access pattern shown in FIG.

  Following the writing of the data of the current command issue number, a SMART information read command is issued, the SMART information is read from the HDD 300, and the number of alternative sectors is extracted from the SMART information.

  Thereafter, the number of alternative sectors before the command issuance number stored in advance in the abnormal value DB recording unit 40 is read and compared with the number of alternative sectors read from the HDD 300, thereby changing the number of alternative sectors in the command issuance number. It is checked whether or not the change exceeds a threshold value to be an abnormal value, and the SMART information read from the HDD 300 is overwritten on the previous SMART information in the abnormal value DB recording unit.

  By repeating this process for the entire HDD, a replacement sector occurrence distribution for the entire HDD is created, and it is confirmed that the error counter does not exceed a specific threshold value that is determined to be abnormal. Failure prediction is performed based on the change in the error counter, and an HDD stop warning is issued when it is predicted that the lifetime has been reached.

  The purpose of the HDD failure prediction apparatus 100 of the present embodiment is to constantly grasp the distribution of alternative sectors on the disk and predict failure. Rather than paying attention to the fact that a substitute sector has occurred in each accessed sector itself, in the area accessed with a specified specific capacity, the percentage of the substitute sector generated relative to the access amount, How the generated areas are distributed on the disk is stored as information.

  If this is performed by a normal file system access, it is necessary to monitor the occurrence of alternative sectors by accessing in units of one sector, and record the information of the sector to be accessed in units of one sector for each access. It is necessary to prepare a large table proportional to

  In addition, since the access length is always variable, a long data length access has a lot of information obtained from the access, so it is easy to predict the factor at the time of occurrence of an abnormality, but a short data length access can be obtained from the access. Since there is little information, even if an abnormality occurs, it is difficult to predict the growth of the failure, and it becomes difficult to make a unified failure prediction judgment for each access.

  In particular, when a short data length is accessed in the same area where a long data length has been accessed due to overwriting, changes in abnormal sectors are monitored only in the short data length area, so the same access Nevertheless, it is not possible to monitor changes in the remaining long data length abnormal sectors accessed previously. For this reason, the occurrence of a replacement sector in this area was seen in the previous access with a long data length, and even if failure growth was observed, the replacement sector in that access was within the normal range by the next short data length access. If considered, it may overlook the growth of disabilities.

  In order to avoid such problems and to accumulate the amount of change of the alternative sector accurately, the capacity (number of sectors) accessed at a time is fixed so that the capacity for monitoring the alternative sector does not change at every access. To do. This fixed capacity is referred to as “specific capacity”. In this embodiment, the specific capacity is 256 sectors, but the present invention is not limited to this, and other sectors may be used. Also, in correspondence with the first LBA (Logical Block Address) used for each access, 1, 2, 3,. . . Those given serial numbers such as are called command issue numbers. In the following description, LBA can be used instead of the command issue number. However, using the command issue number can reduce the data capacity required for storage and computation.

  FIG. 4 is a diagram for explaining an access pattern for a command issue number. As shown in FIG. 4, by accessing in units of specific capacity (256 sectors) and reading information on alternative sectors included in the SMART information immediately after the access, the number of alternative sectors included in the access area of a certain capacity is always generated. Measure the percentage of numbers. A disk area (sector) having a specific capacity accessed for each command issue number is called a “command issue number area”.

  The specific capacity may be the number of sectors corresponding to one track of the disk. The specific capacity may be determined according to the number of sectors on the innermost circumference of the disk.

  By making access in specific capacity units, the size of the monitoring table of alternative sectors can be greatly reduced, and the same amount of access is made to the same area, so the number of alternative sectors in the access area always having the same number of sectors. Therefore, the change of the number of alternative sectors in the specific area can be confirmed efficiently and accurately.

  Since the disk is accessed according to the access pattern of FIG. 4, the access area of the disk corresponding to the command issue number is known for each access. On top of that, you can see which command issue number the alternative sector occurred, so by plotting the change in the number of alternative sectors against the command issue number, how the occurrence of the alternative sector is distributed on the disk I can know. The failure prediction of the HDD is performed by monitoring the change in the distribution of the alternative sector.

  SMART information refers to the progress of the abnormality of the HDD represented by the number of alternative sectors, and the total of actual operation such as the number of OFF / ON, power ON time, seek time, etc. together with the threshold prepared by the HDD manufacturer. It is stored in an internal memory or the like, and is used as an index for HDD failure prediction.

  However, it is not clear how various parameter values of the SMART information are related to the operation of the HDD and in what range, and the parameter values of the SMART information reach the threshold value prepared by the HDD manufacturer. Very often HDD failures occur before. Because it is difficult to predict operational failures simply by relying on the thresholds determined by such manufacturers, the actual situation is that the user has prepared a separate threshold derived from parameter value changes and has performed failure prediction. is there. In this embodiment, failure prediction is performed by evaluating the number of alternative sectors of SMART information by a unique method.

  FIG. 2 is a flowchart showing a failure prediction procedure performed by the HDD failure prediction apparatus 100.

  In step S201 in FIG. 2, it is identified which command issue number corresponds to the top data transferred from the temporary storage unit 20 to the HDD 300. For example, in the temporary storage unit 20, the data to be transferred is managed using the same LBA that is normally written from the host 200 to the HDD 300, and a value obtained by dividing the head LBA by a specific capacity at the time of transfer is calculated and the command issue number and do it. This identified command issue number (first command issue number used for data transfer) is set to i.

  In step S205, it is calculated how many command issue numbers (specific capacities) the data capacity (write capacity) transferred from the temporary storage unit 20 to the HDD 300 corresponds to. Specifically, the data capacity is divided by the specific capacity, and the quotient and remainder are calculated. Then, the quotient is set as the write count M.

  Writing to the HDD 300 can be controlled in units of sectors, but the specific capacity is more than that (here, 256 sectors), so if the last write data is less than the specific capacity, the existing data is in the specific capacity to be written last. May exist.

  Therefore, in step S210, it is determined whether or not the remainder calculated in step S305 is “0”. That is, it is determined whether the write capacity is divisible by the specific capacity. As a result, when it is not divisible (when there is a remainder) (NO in S210), the process proceeds to step S215. In step S215, it is confirmed whether or not there is other data in the (M + 1) th writing area corresponding to the fractional data. If there is other data (YES in S215), the existing data is determined in step S220. After reading the data into the temporary storage unit 20 and combining it with the (M + 1) th write data, the data is prepared as the (M + 1) th write data, and the process proceeds to step S225. If no other data exists in step S215 (NO in S215), it is not necessary to combine the write data, and the process directly proceeds to step S225. As a result, the number of times of writing increases by one per fraction, so M is incremented by 1 in step S225, and the process proceeds to step S230.

  When the write capacity is divisible by the specific capacity in step S210 (YES in S210), the process proceeds to step S230.

  In step S230, the specific capacity corresponding to the command issue number i in the access pattern shown in FIG. 4 is written.

  In step S232, the SMART information after writing to the command issue number i is read from the HDD 300, and the current SMART information is registered in the abnormal value DB recording unit 40. As shown in FIG. 5, the abnormal value DB recording unit stores the SMART information acquired at the previous access and the SMART information acquired at the current access for each command issue number.

  In step S235, the SMART information in the command issue number area is analyzed. The SMART information determination process will be described in detail later with reference to the flowchart of FIG.

  If the cumulative distribution of alternative sector occurrences in the command issue number area analyzed in step S235 exceeds the threshold value, it indicates that a failure is in progress. In step S235, physical errors such as scratches on the board surface are analyzed from the distribution of alternative sector occurrences in the SMART information of each sector so far, and the arrival of the life is comprehensively detected. By issuing a stop warning, the user is prompted to save data from the HDD 300 in operation.

  In step S240, the HDD error counter and SMART information updated in the process of step S235 are displayed. The user (operator) can monitor the status of the HDD 300 by using these counter values, and the operator can judge from this numerical value and stop the HDD 300 independently when necessary.

  In step S245, when it is confirmed that the error counter processed in response to the result of step S235 has exceeded the operation stop parameter value (YES in S245), the failure prediction process ends. Note that immediately after YES in step S245, a warning message that further alerts the user may be displayed, or a message notifying that the failure prediction process is to be terminated may be displayed.

  If NO in step S245, 1 is subtracted from the write count M in step S250, and 1 is added to the command issue area i because the access moves to the next area.

  Finally, in step S255, when the number M of writing is larger than 0 (YES in S255), a series of processing from step S230 to step S250 is repeated until the predetermined number of writings is reached. When the number M of times of writing becomes 0 (NO in S255), the failure prediction process is terminated.

  FIG. 3 is a flowchart showing a detailed procedure of the SMART information determination process in step S235. In the SMART information determination process, the state of the sector at the command issue number is analyzed. In order to analyze the state of the sector, paying attention to the change in the number of alternative sectors in the SMART information, the failure prediction of the HDD 300 is finally performed from the change in the number of alternative sectors.

  The SMART information determination process is performed by adding the error counter value for each command issue number shown in FIG.

  Command issue numbers are grouped for every predetermined number N of command issue numbers. Hereinafter, this group is referred to as “command issue segment” or simply “segment”. The predetermined number N is referred to as “segment length”. Typically, N = 30-50 is good. For example, when N = 30, command issue number = 1-30 is segment 1, command issue number = 31-60 is segment 2, command issue number = 61-90 is segment 3, and so on. And the segment.

  In step S300, the SMART information (particularly the number of alternative sectors) when the previous command issue number i was executed is read from the abnormal value DB recording unit 40.

  In step S305, the number of alternative sectors of the previous SMART information read in step S300 is compared with the number of alternative sectors of the current SMART information read in step S232 of FIG. At this time, the maximum number of alternative sectors that may occur is the number of sectors in the command issue number.

  If the number of current alternative sectors is greater than the previous number of alternative sectors in step S305, that is, if the occurrence of a new alternative sector is confirmed (YES in S305), the current command issue number area is set in step S310. Check whether the number of alternative sectors has increased by a predetermined number (for example, 5) or more (whether a predetermined number of alternative sectors have been newly generated) or not. Check for an increase.

  When scratching of the head due to impact occurs, the number of alternative sectors increases rapidly for consecutive sectors on the same circumference of the board surface of the HDD 300 due to the scratch. For this reason, in step S310, it is confirmed whether or not an alternative sector is suddenly generated only by access using the current command issue number. In the case of a failure due to a flaw, the function of the HDD 300 is temporarily recovered by the occurrence of an alternative sector, but the failure is likely to extend beyond the current command issue number, so the HDD is stopped immediately, You need to take measures to protect your data. If it is confirmed in step S310 that an alternative sector equal to or more than a preset threshold value (for example, 5) is generated (YES in S310), it is necessary to issue an HDD stop recommendation in step S330 to stop the use of the HDD 300.

  It is desirable to change the threshold value for the number of alternative sectors generated for each HDD 300 so that the smallest scratch can be detected early on the innermost circumference of the disk with the fewest consecutive sectors.

  FIG. 6 shows the number of alternative sectors generated in the current command issue number area in step S315 when the occurrence of an abrupt alternative sector exceeding the threshold cannot be confirmed in the current command issue number area (NO in S310). Register in the error counter.

  In step S320, when a total of five or more non-zero error counters are confirmed in the adjacent segment area of 10 segments before the access target segment and 10 segments after the access target segment (YES in S320), a delay occurs in the adjacent segment area. The process proceeds to step S330 and issues a HDD stop warning. This is because if a plurality of alternative sectors are included in the neighboring segment area, it indicates that the failure in the command issue number area spreads around the current command issue number area with high density.

  As described above, the failure occurrence position is evaluated by referring to the SMART information of the HDD 300 in the neighboring segment area, thereby grasping the progress of the failure and predicting the failure.

  The number of adjacent neighboring segment areas, 10 before and 10 after the segment to be accessed, and the threshold of 5 or more error counters in the neighboring segment area indicate that the distribution of the occurrence of alternative sectors that can predict failures Although it is specified as a value that is clearly observed, it is a typical example, and other values may be used. The values of these parameters vary with the capacity of the HDD 300. When the maximum capacity of the HDD 300 is small, since the command issue number area for accessing the entire HDD 300 is small, the delayed access area easily spreads with a slight change. Therefore, it is necessary to reduce these parameter values. When the maximum capacity of the HDD 300 is large, since the command issue number area for accessing the entire HDD 300 is large, the delayed access area is difficult to spread with a slight change, so these parameter values need to be increased.

  If a failure spread is detected, it is likely that the physical failure is not limited to the command issue number area currently being evaluated, but is also spread to the command issue number region that has not yet been evaluated. Since there is a possibility that data may not be read due to development up to a read error, it is necessary to promptly issue an HDD stop recommendation and stop using the HDD.

  As described above, the progress of the failure can be analyzed from the failure analysis result, and the error counter is normally updated unless the data needs to be taken out from the HDD immediately. In step S325, the error counters of all segments are a predetermined number ( For example, if it exceeds 10 (YES in S325), an HDD stop warning is issued in step S330.

  The threshold that the error counter is 10 or more throughout all segments is that the failure progresses when the alternative sector and the access delay area increase on the HDD 300 even when the failure does not show a specific spread. It is a value indicating that The threshold value of 10 indicates the total number of values that start to cause a host access failure with the number of occurrences per capacity of the HDD 300, and this value varies with the capacity of the HDD 300. Other numbers may be used as the predetermined number.

  As described above, according to the failure prediction procedure performed by the HDD failure prediction apparatus 100 according to the present embodiment, the failure occurring from the occurrence position and the number of occurrences of the alternative sector by clarifying the occurrence position of the alternative sector. The severity can be evaluated by weighting and the progress of the failure can be accurately grasped. As a result, failure of the HDD 300 can be accurately predicted, and loss of data in the HDD 300 can be prevented.

  If the function is restored due to the occurrence of a substitute sector, if it is a single occurrence, the error count will not increase rapidly as the HDD 300 function has been normalized, but the error count will increase in a contiguous area of neighboring segments. If this is the case, it is considered that a flaw due to the occurrence of an alternative sector or a write error due to a defective head has occurred. In this way, by grasping the number of occurrences of the alternative sector of SMART information in each segment as an error count, the distribution of the progress of the segment in which the problem has occurred can be clarified in the entire disk. From the progress state, it can be determined that the failure of the HDD 300 is near.

  Finally, the error counter is used to determine from the counter value accumulated due to the change in the number of alternative sectors that the threshold for failure prediction has been exceeded and to notify the HDD 300 of a stop warning by an indicator lamp or the like. This may be a warning by a buzzer or the like, and may be linked with a stop function of the present system. In addition, in cooperation with an external system, according to the value of the error counter, information related to the purchase of a storage device such as an HDD (advertising information, etc.) is displayed, the use of a backup service using the cloud or the like is promoted, A preferential service (such as presentation of a coupon) for promoting the purchase of a storage device or the use of a backup service may be implemented. In this way, by providing pinpoint appropriate information to highly necessary users, it is possible to improve user convenience and increase sales of related products and services.

(Embodiment 2)
When an HDD malfunctions, there is no guarantee of written data. Once a malfunction of the HDD occurs, internal data cannot be read basically, resulting in a large loss. That is, at the stage where it is noticed that there is a failure, it is impossible to avoid losing some data in the HDD. Therefore, the HDD failure prediction apparatus 100 according to the embodiment of the present invention predicts the life of the HDD as a limit at which the operational data can be normally read, and avoids losing data due to the HDD failure. With the goal.

  Japanese Patent Laid-Open No. 2011-68109 describes a method for evaluating access time for HDD life prediction by comparing transfer times. However, with regard to HDD failures, failures that come from wear of internal parts due to deterioration over time, in particular, tend not to stop at the location where the failure has occurred, but to expand over time, so the signs cannot be captured with high accuracy. As the time elapses, the possibility of losing the internal data of the HDD increases.

  In particular, re-accessing an area that is progressing to a failure is likely to progress to a serious failure, so the ability to predict the progression of the failure will determine how to take subsequent HDD internal data remedies. This is a very important guideline.

  Therefore, in the HDD failure prediction apparatus 100 according to the second embodiment, the HDD SMART information that is not used in Japanese Patent Laid-Open No. 2011-68109 is used, and based on the scatter diagram of the access time according to a specific access pattern, A method of increasing the accuracy of failure prediction by determining a threshold value used when predicting failure according to access time is adopted.

  The configuration diagram of the HDD failure prediction apparatus 100 according to the second embodiment is the same as the configuration diagram of the HDD failure prediction apparatus 100 according to the first embodiment shown in FIG. Here, description of the configuration and operation common to the first embodiment will be omitted as appropriate, and configuration and operation different from the first embodiment will be described.

  The control unit 30 measures the command execution time (access time) at the time of writing by the HDD controller 10, extracts the number of alternative sectors from the SMART information read by the HDD controller 10, and changes the access time and the number of alternative sectors. From the above, the life of the HDD 300 is determined.

  The abnormal value DB recording unit 40 stores a threshold value for determining an abnormality in access time, and uses as a database the command issue number at which the delayed access occurred when the delayed access occurs, the delayed access time, and the SMART information when the delayed access occurs. sign up. In this embodiment, as will be described later, the HDD 300 is accessed in a predetermined number of sectors (specific capacity units).

  When the host 200 issues a data write command to the HDD 300, the written data is temporarily stored in the temporary storage unit 20 via the HDD controller 10. When the temporarily stored write data reaches a predetermined capacity, the HDD controller 10 writes to the HDD 300 from the old data on the time axis according to the access pattern shown in FIG. Details of this processing will be described later.

  At this time, the control unit 30 measures the command execution time (access time) when data is written from the HDD controller 10 to the HDD 300, reads the threshold value stored in advance in the abnormal value DB storage unit, and reads the current access time. It is determined whether or not exceeds this threshold.

  When it is confirmed that the access time exceeds the threshold, the control unit 30 stores the LBA, the access time and the SMART information at that time when the access time exceeds the threshold, and a warning (warning) counter and an error counter described later. The failure of the HDD 300 is predicted based on the transition of the warning counter and the error counter. When it is predicted that the life has been reached, an HDD stop warning is issued.

  The number of alternative sectors in the SMART information is an important value for knowing the failure of the HDD, but the number of alternative sectors in the HDD SMART information simply indicates the number of occurrences, and the sector that is the alternative sector There was no way to know in real time during the operation of the HDD. For this reason, SMART information is mostly used for analysis of causes after an internal failure of the HDD occurs in existing technologies.

  In the actual operation, the alternative sector does not occur suddenly, and an alternative sector is generated when writing to the sector originally intended to be written after a predetermined number of retries by the HDD manufacturer is not possible. When a sector in which an alternative sector has occurred is accessed in the process of failure progression, a retry occurs as a sign of failure, resulting in an access delay. In this retry process, if the data is read and an alternative sector is generated, the data is protected. However, in the process of repeating the retry, the data cannot be read and may not be transferred to the alternative sector. If the failure progresses so far, the data written in the sector where the failure has occurred will be lost.

  By identifying and counting the sector where the failure has occurred in the operating HDD, it is possible to know in which sector the failure is progressing over time, and from the failure distribution count, It is possible to predict the arrival.

  Therefore, in the HDD failure prediction apparatus 100 of the second embodiment, the access time threshold value during normal operation of the HDD determined in advance by a specific method is recorded in the abnormal value DB recording unit 40, and the HDD access time is recorded. The SMART information typified by the number of alternative sectors is read as a trigger when the threshold value during normal operation is exceeded, and compared with the SMART information when the access time exceeds the threshold value last time. Basically, a sector that requires an access time exceeding the threshold value takes extra processing time due to an access abnormality, so the control system inside the HDD may leave the abnormality content as SMART information. Extremely expensive. However, even if it takes an access time exceeding the threshold value in this way, if data is written or read from the sector, it may not be reflected in the SMART information.

  As described above, when the access time exceeds the threshold value, it is indicated that a factor causing some access delay is generated in the HDD as compared with the normal time. However, information such as an access time threshold value that leads to the update of SMART information differs for each HDD manufacturer and is not always clear. However, if SMART information is read in advance with reference to the threshold of access time during normal operation of the HDD, threshold data for failure determination is obtained from the HDD manufacturer, or complex data analysis such as prior failure analysis is performed. There is an advantage that HDD failure can be predicted without having to

  As described above, the HDD failure prediction apparatus 100 according to the second embodiment determines in which sector the SMART information changes when the access time exceeds the threshold, based on the access time threshold during normal operation of the HDD. By grasping and tabulating, the progress of failures inside the HDD can be grasped more accurately. Since the SMART information and the state of the sector are correlated in real time to detect a sign of a failure occurrence, the failure can be predicted with high accuracy.

  Here, an outline of the access time of the HDD 300 will be described. FIG. 14A shows the access time when the HDD 300 is normal, and FIG. 14B shows the access time when an abnormality occurs in the HDD 300.

  As shown in FIG. 14A, the normal access time can be represented by the sum of the data transfer time and the time depending on the head operation such as seek time, rotation waiting time, and focusing time.

  As shown in FIG. 14 (b), the access time when an abnormality occurs is several tens of times that during normal access because a retry time, alternative sector processing time, etc. are added in addition to the normal access time. Processing time is required. Therefore, there is a relation between the access time and the change of the alternative sector, and in the sector where the access time is extended, the alternative sector is generated or is highly likely to occur in the SMART information read immediately after this.

  Since the disk rotates in the HDD 300 at a constant speed, when the head accesses the target sector, a rotation waiting time corresponding to one rotation of the disk occurs at maximum depending on the access timing. In addition, since the seek time from the next access position varies depending on the head position when the previous access is completed, the total access time also varies. As a result, since the convergence time until access is different, the target sector cannot be accessed in the same access time as at the time of factory shipment, and varies for each access. Therefore, in the present embodiment, variations in access time are dealt with by using the access pattern shown in FIG.

  FIG. 7 is a flowchart showing a failure prediction procedure performed by the HDD failure prediction apparatus 100.

  In step S401, the HDD controller 10 accesses the HDD 300 according to the access pattern of FIG. 11, and the control unit 30 measures the command execution time at that time.

  In the access pattern shown in FIG. 11, data of a specific capacity is written in correspondence with one command issue number. In this embodiment, the specific capacity is 256 sectors, but the number of other sectors may be used and is not limited to this. In the case of the HDD 300, if the head position at the time when the previous access is ended is unspecified, the seek time varies particularly, and the relatively accurate access time cannot be measured. Therefore, as shown in FIG. 11, the head position is always reset to the initial position (in this case, the position of sector 0) after the end of access, and then the specific capacity (256 sectors) is sequentially written, thereby enabling a more accurate access time. Measurement is possible.

  FIG. 12A is a schematic diagram of access time when the HDD 300 is accessed according to the access pattern of FIG. The horizontal axis represents the command issue number, and the vertical axis represents the access time for each command issue number. FIG. 12B is a graph showing the schematic diagram of FIG. 12A with actually measured values. When the measurement data is plotted, a scatter diagram as shown in FIG. 12B is obtained, and the access time is distributed in the shape of a band rising upward.

  If the HDD 300 is in an ideal state without waiting for rotation and focusing time, the scatter diagram plotting the access time for the sector to be the access target should be almost one line. However, as described above, since there is always a variation in the waiting time for rotation and the focusing time for disk access, plotting the access time for each specific capacity shown in the access pattern of FIG. 11 actually specifies it as shown in FIG. 12B. There is a strong correlation between the command issue number and the access time. The inclination of the entire band in FIG. 12B is mainly due to an increase in seek time due to the fact that the sector to be accessed becomes far from the head reset position as the command issue number increases as shown in FIG.

  Note that the command execution time (access time) may be measured by the HDD failure prediction apparatus 100 or by an external device that generates the same access pattern.

  In step S402, a value corresponding to the upper end of the band in the scatter diagram measured in step S401 is detected. This value is used as a threshold value (failure prediction threshold value) for each command issue number. This value is the access time when the waiting time depending on the head operation such as seek time, rotation waiting time, and focusing time is maximized. When the HDD 300 is normal, the access time is less than this value. There are features. Therefore, by using this value as a failure prediction threshold, it is possible to grasp that the access of the HDD 300 is not normal when the access time exceeds this threshold. A specific method for detecting this threshold will be described later.

  In step S403, the threshold value calculated in step S402 is registered in the abnormal value DB recording unit 40. Specifically, as shown in FIG. 15A, a command issue number and a threshold value are recorded in association with each other.

  Steps S401 to S403 are pre-processing or initial setting processing for performing failure prediction.

  In step S404, a failure prediction operation is performed when the HDD 300 is used. The control unit 30 monitors whether or not the access time of the command issue area exceeds the threshold value recorded in the abnormal value DB recording unit 40, and for the command issue number area exceeding the threshold value, the command issue number is stored in the abnormal value DB recording unit 40. And the access time and SMART information are recorded. The current SMART information is compared with the SMART information when the access time exceeded the threshold value last time. Therefore, the abnormal value DB recording unit 40 temporarily stores the SMART information of the command issue number when the access time exceeds the threshold value last time, and uses it for comparison with the SMART information of the current command issue number.

  The control unit 30 counts the warning counter and the error counter recorded in the abnormal value DB recording unit 40. As a result, the sector in which the access time in the HDD exceeds the threshold and the distribution of changes in the SMART information in the sector are predicted to fail. Is reached, the HDD 300 is issued a stop warning and the process is terminated.

  In the processing of steps S401 to S403, the control unit 30 operates as a threshold value calculation unit. The threshold calculation unit may be a circuit different from the control unit 30.

  Note that if the HDD model number and firmware are the same, the threshold value obtained from the band of the scatter diagram in FIG. 12B is the same, so there is no need to create a new threshold value, so steps S401 to S403 are omitted. The failure prediction in step S404 can be started using the threshold value measured by another HDD. When using a threshold value measured with another HDD, the HDD failure prediction apparatus 100 does not need to include a threshold value calculation unit.

  Here, the method for detecting the threshold value in step S402 will be described in detail. The first method in step S402 will be described.

  As shown in FIG. 13, the command issue numbers are grouped for every predetermined number N of command issue numbers. Hereinafter, this group is referred to as “command issue segment” or simply “segment”. The predetermined number N is referred to as “segment length”. Here, the predetermined number N (segment length) is set so that one segment includes approximately one or more points at which the waiting time depending on the head operation such as seek time, rotation waiting time, and focusing time is maximized. To do. Typically, N = 30-50 is good. For example, when N = 30, command issue number = 1-30 is segment 1, command issue number = 31-60 is segment 2, command issue number = 61-90 is segment 3, and so on. And the segment.

  Next, the maximum value of the access time is detected for each segment. The maximum value is set as a threshold value in the segment. For example, when N = 30 and the command issue number = 12, the access time is maximum and the maximum value is 30 msec in the segment 1, all the threshold values corresponding to the command issue numbers 1 to 30 are set to 30 msec. . Alternatively, a value obtained by multiplying the maximum value of the access time in each segment by a predetermined magnification may be used as the threshold of the segment. For example, the predetermined magnification = 1.2 and the maximum value 30 msec × 1.2 = 36 msec may be set as the threshold value of the segment. Or it is good also considering the value which added predetermined value to the maximum value of the access time in each segment as a threshold value. For example, the predetermined value = 5 msec and the maximum value 30 msec + 5 msec = 35 msec may be set as the threshold value.

  The threshold values calculated by the first method are all the same for the command issue numbers corresponding to one segment. Therefore, in step S403, the segment number and the threshold value may be recorded in correspondence with each other as shown in FIG. 15B without recording the threshold value for each command issue number.

  The second method of step S402 will be described. First, as in the first method, a segment is formed for each predetermined number of command issue numbers. This segment is for use in subsequent processing steps, and in step S402, no segment is used.

  Next, for a certain command issue number (command issue number i), the maximum value of the access time is detected for command issue numbers (i-w to i + w) in a predetermined range before and after the command issue number. That is, the threshold value θ [i] corresponding to the command issue number i is calculated according to Equation (1). Here, a [i] is the command execution time (access time) corresponding to the command issue number i, w is a positive integer, and max is a function that returns the maximum value from the values specified in the arguments. is there. According to Equation (1), the maximum value is detected for (2w + 1) command issue numbers. The positive integer w may be set so that one or more maximum values of access time are included in (2w + 1) command issue numbers. Typically, w = 15-25 may be used. (2w + 1) may be the same as or different from the segment length N. Assuming that the maximum command issuance number used for measurement in step S401 is P, threshold values θ [w + 1] to θ [Pw] corresponding to i = (w + 1) to (P−w) are calculated according to mathematical formulas. . For i = 1 to w, θ [w + 1] may be used, and for i = (P−w + 1) to P, θ [Pw] may be used.

  In addition, the threshold may be calculated by further performing a moving average process on the value calculated according to Equation (1). For example, the left side of Expression (1) is substituted into the temporary variable μ [i], and the moving average of μ [i] is calculated according to Expression (2) as the threshold θ [i]. Here, ε [j] is a weighting coefficient that satisfies Equation (3). L is a positive integer and is typically set to 5-10. By calculating the threshold value according to the mathematical formula (2), the change in the threshold value becomes smooth and failure prediction may be possible with high accuracy.

  FIG. 8 is a flowchart showing a detailed procedure of the failure prediction process shown in step S404.

  The failure of the HDD 300 indicates how much the access time of the area delimited by the specific capacity unit exceeds the threshold which is the upper end of the band of the scatter diagram shown in FIG. 12B and how the access delay occurs. Predict based on whether it spreads over again.

  The cause of the access delay exceeding the threshold value of the HDD 300 is that an abnormality has occurred at the time of access, and in addition to the processing at the time of normal access, the processing time at the time of occurrence of an abnormality such as retry or occurrence of alternative sector is added.

  However, the occurrence of a replacement sector due to damage to a specific sector causes a large access delay only when the replacement sector occurs. However, even if the same sector is accessed thereafter, the HDD 300 is treated as having returned to normal operation. However, even if the same area is accessed again, the access delay does not occur. In this case, the occurrence of the abnormal sector is not confirmed, and the expansion of the abnormal sector cannot be confirmed, and the expansion beyond the alternative sector generated at that time is not seen. On the other hand, the access delay due to deterioration over time gradually increases as the deterioration progresses, and the number exceeding the threshold of the area delimited by the segment length increases for each access, but the pattern cannot be specified, so it is delimited by segment. A failure is predicted by registering in the abnormal value DB recording unit 40 the transition of how the threshold of the specified area is exceeded.

  In step S501 of FIG. 8, it is identified which command issue number corresponds to the top data transferred from the temporary storage unit 20 to the HDD 300. For example, in the temporary storage unit 20, the data to be transferred is managed using the same LBA that is normally written from the host 200 to the HDD 300, and a value obtained by dividing the head LBA by a specific capacity at the time of transfer is calculated and the command issue number and do it. This identified command issue number (first command issue number used for data transfer) is set to i.

  In step S505, it is calculated how many command issue numbers (specific capacities) the data capacity (write capacity) transferred from the temporary storage unit 20 to the HDD 300 corresponds to. Specifically, the data capacity is divided by the specific capacity, and the quotient and remainder are calculated. Then, the quotient is set as the write count M.

  Writing to the HDD 300 can be controlled in units of sectors, but the specific capacity is more than that (here, 256 sectors), so if the last write data is less than the specific capacity, the existing data is in the specific capacity to be written last. May exist.

  Therefore, in step S510, it is determined whether or not the remainder calculated in step S505 is “0”. That is, it is determined whether the write capacity is divisible by the specific capacity. As a result, when it is not divisible (when there is a remainder) (NO in S510), the process proceeds to step S515. In step S515, it is checked whether there is other data in the M + 1th writing area corresponding to the fractional data. If there is other data (YES in S515), the existing data is temporarily stored in step S520. After reading into the storage unit 20 and combining with the (M + 1) th write data, the data is prepared as the (M + 1) th write data, and the process proceeds to step S525. If no other data exists in step S515 (NO in S515), it is not necessary to combine the write data, and the process directly proceeds to step S525. As a result, the number of times of writing is increased by one per fraction, so M is added by 1 in step S525, and the process proceeds to step S530.

  If the write capacity is divisible by the specific capacity in step S510 (YES in S510), the process proceeds to step S530.

  In step S530, the specific capacity corresponding to the command issue number i in the access pattern shown in FIG. 11 is written.

  In step S535, it is determined whether or not the command execution time a [i] corresponding to the command issue number i has exceeded the threshold value θ [i].

  As a feature with respect to the access time of the HDD 300, the HDD 300 in which failure and deterioration have not progressed finishes within the normal processing time, so the segment access time falls within the threshold. However, the HDD 300 that has deteriorated over time and the HDD 300 in which a failure has occurred have increased internal failures due to head contamination, the effect of accumulated dust, scratches on the board surface, and the like. Since time requires processing corresponding to a failure and time is required for internal processing compared to normal processing, the access time exceeds a predetermined threshold value, and the delay of the access time for the write address increases. In step S535, it is confirmed whether or not such a symptom has occurred.

  When the command execution time does not exceed the threshold (NO in S535), the process proceeds to step S545. If past warnings corresponding to the command issue number have been counted in step S545 (YES in S545), it is considered that there is a temporary factor at the time of access, and the command issue number area warning count is set to 0 in step S550. Return to step S565.

  When the command execution time exceeds the threshold (YES in S535), in order to confirm whether the access delay time changes the SMART information, the SMART information for the command issue number i is read in step S540, and the abnormal value DB is recorded. The SMART information is recorded in the unit 40.

  FIG. 19 shows a memory structure for storing SMART information in the abnormal value DB recording unit 40. The storage memory A always stores the SMART information that is being processed, and the storage memory B stores the SMART information that was previously processed.

  Specifically, when the SMART information is read in step S540, the data in the storage memory A (SMART information read last time) is first copied to the storage memory B. The data originally stored in the storage memory B (SMART information read last time) is overwritten. Thereafter, the SMART information read this time is stored in the storage memory A.

  Thereafter, in step S543, the current SMART information in the storage memory A is compared with the SMART information that has been subjected to the previous process in the storage memory B, and it is determined whether or not the SMART information (here, the number of alternative sectors) has been updated. If there is no update (NO in S543), a warning determination process in step S555 is performed. If there is an update (YES in S543), an error determination process in step S560 is performed.

  Here, when the SMART information is updated, it indicates that the SMART information has been updated in the sector within the command issue number accessed now. When the access time exceeds the threshold for each command issue number, grasp the change in SMART information and count the warning counter and error counter for each specific segment to determine how the sectors judged abnormal are on the HDD. It is analyzed whether it is distributed to.

  In step S555, a warning determination process for detecting a non-fatal failure is executed. If the failure is such that the SMART information is not updated while the access time exceeds the threshold, or if the SMART information is updated other than the number of alternative sectors in step S543, it is determined that the failure is not fatal. Since there are those that have detected the occurrence of, warning processing is performed and the warning counter is counted. The warning determination process will be described in detail later with reference to the flowchart of FIG.

  In step S560, an error determination process for detecting a fatal failure is executed. In step S560, attention is paid to each sector included in the command issue number area, the actual state of the access time delay is grasped, and the state of each sector is analyzed mainly based on the SMART information acquired in real time. In particular, a physical error such as a scratch on the board surface is analyzed from the distribution of the occurrence of alternative sectors in each sector, and the arrival of the life is comprehensively detected including the analysis result of the warning determination process in step S555. By issuing a warning to stop the HDD, it is urged to save data from the HDD in operation. The error determination process will be described in detail later with reference to the flowchart of FIG.

  In step S565, the warning counter value and error counter value of the HDD 300 updated in the processes in steps S555 and S560 are displayed together with the SMART information. The user (operator) can monitor the status of the HDD 300 by using these counter values, and the operator can judge from this numerical value and stop the HDD 300 independently when necessary.

  If it is confirmed in step S570 that the error counter processed in response to the result of step S560 has exceeded the operation stop parameter value (YES in S570), the failure prediction process is terminated. Note that immediately after YES in step S570, a warning message that further alerts the user may be displayed, or a message notifying that the failure prediction process is to be terminated may be displayed.

  While data is being written to the HDD 300, the processes in steps S555 and S560 are performed on the command issue number to predict failure of the HDD 300 being used. For this reason, 1 is subtracted from the write count M in step S575, and 1 is added to the command issue area i because the access moves to the next area.

  Finally, in step S580, if the number of times of writing M is greater than 0 (YES in S580), a series of processing from step S530 to step S575 is repeated until the predetermined number of times of writing is reached. When the number M of times of writing becomes 0 (NO in S580), the failure prediction process is terminated.

  FIG. 9 is a flowchart showing a detailed procedure of the warning determination process in step S555. In the warning determination process, an area in which an abnormality that cannot be detected by a change in the number of alternative sectors in the SMART information is identified in the HDD, and it is determined how the abnormality occurrence area is distributed (spread).

  The warning determination is performed by adding the value of the warning counter for each command issue number shown in FIG.

  Since the warning determination process is performed when the access time of the accessed command issue number exceeds the threshold, in step S601, the warning counter is incremented by one, and the command issue number and the command execution time (access time) at that time are set. Register in the abnormal value DB recording unit 40.

  In step S605, in order to reliably detect the spread of the failure, in the narrow neighboring segment area (first neighboring segment area) in which five segments are taken before the access target segment and five segments are taken later, the first neighboring segment area shown in FIG. Check if the scatter plot band is spreading beyond the threshold. Specifically, it is determined whether 10 or more non-zero warning counters have occurred in the first neighboring segment area. If 10 or more warning counters have been confirmed (YES in S605), it is determined that the delay has increased in this neighboring segment area, and a warning for stopping the HDD 300 can be issued in the next error determination process. In S610, a failure prediction threshold value (here, 10) is added to the error counter.

  The warning determination process is basically performed when the access time exceeds the threshold value as shown in step S535 of FIG. 8 and an access failure has occurred, but in step S543, the occurrence of an alternative sector cannot be confirmed in the SMART information. Is called.

  In the update of the SMART information, particularly the alternative sector, it is possible to recognize that a failure has occurred in the internal system of the HDD, and it is possible to determine its own abnormality by counting the number of alternative sectors. Faults that can be physically determined such as scratches due to head contact can be determined by increasing the number of alternative sectors. However, in life prediction, access delays occur once, and there are no alternative sectors in the neighboring sectors centered on some sectors where access time does not recover, but access exceeding the threshold There is a feature that the time delay gradually increases and the alternative sector suddenly expands from a certain time. Therefore, it is necessary to record the distribution of access delay for each sector and determine how the sector exceeds the threshold in the scatter diagram band of FIG.

  Therefore, in addition to the fact that the access time of the command issue number i exceeds the threshold and no alternative sector has occurred, it is determined whether the similar behavior is seen in the sector near the command issue number i, Predict. This is because the access delay of each sector is not so great, but many access delays occur in the area near the command issue number i. It means that some kind of obstacle that extends the access time is in progress. In step S605, an area segmented by the same number of sectors is used as a neighboring segment area so that accurate determination can be made using the number of sectors (command issue numbers) in which access delays have occurred in the neighboring areas. In step S605, the predetermined number such as five segments before, five segments after, and 10 or more warning counters is an example, and a predetermined number other than the above-described values may be used.

  In step S615, when N command issuance numbers in the segment to be accessed are viewed, if there is a predetermined number (for example, 2) of non-zero warning counters, a failure that cannot be recovered in the segment has occurred. In step S620, the error count is incremented by 1, and the error counter value is passed to the error determination process. At this time, since the area where the error has occurred can be specified, it is possible to extend the life of the entire HDD by taking measures such as avoiding the use of the problem area.

  As described above, in the warning determination process, it is possible to determine that a failure that is not counted as an error is accumulated in the HDD 300 by adding the warning counter. Further, when a predetermined number or more of warning counters are generated in a certain segment area, it is possible to immediately connect to an error determination process by adding an error counter.

  FIG. 10 is a flowchart showing a detailed procedure of the error determination process in step S560. In the error determination process, occurrences of alternative sectors in the SMART information are totaled, and the state of the sector in the command issue number area where the alternative sectors are generated is mainly analyzed.

  Pay particular attention to the change in the number of alternative sectors in the SMART information in order to analyze the state of the sector. From the relationship between the generation position of this alternative sector and the command issue number area in which the access time exceeds the threshold, many access delays have already occurred. It is analyzed whether an alternative sector has occurred in the segment of the generated command issue number, or whether an alternative sector has occurred in an area where no access delay has occurred so far. As in the former case, if an alternative sector has occurred within the command issue number included in the segment whose access time already exceeds the threshold, it is determined that a failure has been reached, and the HDD failure prediction is finally made. I do. Like the latter, the occurrence of a single alternative sector is expected to have caused an accidental sector destruction, as seen in the initial failure of the sector itself. Can continue to use normally.

  The generation of an alternative sector is a function recovery method for truncating a sector in which an unrecoverable failure has occurred, switching to a new alternative sector, and returning the HDD to normal operation. Since the generation of the alternative sector itself is not bad, by analyzing the relationship between the generation position of the alternative sector and the command issuance number that is delayed in access, it is determined whether the generation of the alternative sector is a failure. To do.

  In the error determination process, the command issue numbers whose access times exceeded the threshold in the warning determination process are totaled, and when the warning counter exceeds a certain value in the area, it is added as an error count. In step S543 of FIG. 8, when the number of alternative sectors is updated as compared with the previous SMART information, the warning determination process is skipped and the error determination process is performed. This is because the occurrence of an alternative sector means that an abnormality in the HDD has already reached a level that can be recognized by the HDD itself, and it is necessary to determine whether the HDD has stopped, etc. as soon as possible. This is because the command issuance numbers exceeding the threshold, which can be determined as failure predictions, should be tabulated so far.

  The error determination process is performed by adding the error counter value for each command issue number shown in FIG.

  In step S710, as error determination, attention is first paid to only the alternative sector, and it is confirmed whether the alternative sector is abruptly generated although the access delay of the command issue number does not occur so much. For example, if the head is scratched due to an impact, the number of alternative sectors increases rapidly for consecutive sectors on the same circumference of the disk surface of the HDD. In this way, it is confirmed whether an alternative sector is suddenly generated only by accessing with the current command issue number.

  In the case of a failure due to a flaw, the HDD function is temporarily recovered due to the occurrence of an alternative sector, but the failure is likely to extend beyond the current command issue number, so the HDD is stopped immediately, You need to take measures to protect your data. Therefore, when it is confirmed in step S710 that five or more alternative sectors are newly generated (YES in S710), the process proceeds to step S730 to issue an HDD stop recommendation and stop the use of the HDD.

  Here, it is desirable to change the threshold value for the number of alternative sectors generated for each HDD so that the smallest scratch can be found at an early stage on the innermost circumference of the disk with the fewest consecutive sectors.

  In step S710, if it is not possible to confirm the occurrence of 5 or more abrupt alternative sectors in the current command issue number area (NO in S710), the process proceeds to step S715, where an alternative sector is generated in the current command issue number area. Add the number to the error counter. Here, the error counter may have already been added in the warning determination process of FIG. 9, and is further added and accumulated depending on the number of alternative sectors generated in the error determination process.

  Next, in step S720, when five or more non-zero error counters are confirmed in the second neighboring segment area in which 10 segments are taken before the access target segment and 10 segments are taken later (YES in S720). Then, it is determined that the access delay is extended in the second neighboring segment area, and the process proceeds to step S730 to issue an HDD stop warning. This means that alternative sectors are generated in the segments that have been counted in the warning determination process and whose access time exceeds the threshold. In such a case, it is highly possible that the physical failure stays in the command issue number area that is currently being evaluated, and that it has spread to the command issue number area that has not yet been evaluated. Since there is a possibility that the data cannot be read out, it is necessary to immediately issue an HDD stop warning and stop the use of the HDD.

  In this way, it is confirmed that the access delay exceeds the threshold value of the upper end of the band in FIG. 12B due to the spread of the failure in a wide range in the first neighboring segment, and the failure spreads. By identifying the location of the failure in the extended area based on the SMART information of the HDD, the progress of the failure is accurately captured and the failure is predicted.

  The number of adjacent second neighboring segment areas that are 10 before the access target segment and 10 after that, and the number of error counters within the second neighboring segment area are 5 or more are examples, and other than this. The value of may be used. These numbers vary with the capacity of the HDD. Since the command issue number area for accessing the entire HDD is small when the maximum capacity of the HDD is small, the band of the scatter diagram of FIG. 12B spreads beyond the threshold value with a slight change, so these values need to be reduced. Since the command issue number area for accessing the entire HDD is large when the maximum capacity of the HDD is large, the band does not spread with a slight change, so these values need to be increased.

  In step S720, when the number of non-zero error counters in the second neighboring segment area is less than 5 (NO in S720), the process proceeds to step S725. If the total value of error counters in all segments is equal to or greater than a predetermined value (for example, 10) in step S725 (YES in S725), an HDD stop warning is issued in step S730.

  The total value of error counter values in all segments is 10 (predetermined value) or more, even if there is no relationship with the total of access time exceeding the threshold, the alternative sector or access delay area on the HDD Are scattered and indicate an increase in disability. Here, the number of 10 is merely an example, but indicates the total number of values that start to cause a host access failure by the number of occurrences per HDD capacity, and this value varies with the capacity of the HDD. Instead of the condition that the total value of the error counter values in all segments is equal to or greater than a predetermined value, a condition that the number of error counters having a value greater than 0 in all segments is equal to or greater than a predetermined value may be used.

  As described above, the progress of the failure can be analyzed from the failure analysis result, and unless the data needs to be taken out from the HDD immediately, the error counter is added in the warning determination process or the error counter is added based on the number of alternative sectors generated in step S725. When the error counter exceeds 10, an HDD stop warning is issued in step S730.

  As described above, according to the failure prediction procedure performed by the HDD failure prediction apparatus 100 according to the second embodiment, since the SMART information is read only when the access time exceeds the threshold, it is necessary to issue an unnecessary SMART information read command. There is no. Moreover, since the abnormality is detected by combining the threshold comparison of the access time and the change of the SMART information (particularly the number of alternative sectors), the accuracy of failure prediction is increased.

  Even when the SMART information is not updated, the warning determination process and the error determination process are performed. Therefore, it is possible to cope with a minor abnormality in which the SMART information is not updated, and the accuracy of failure prediction is improved. In addition, when the SMART information is updated, the warning determination process is skipped and the error determination process is performed, so that a severe abnormality can be detected efficiently.

  Further, failure prediction of the HDD 300 can be performed with high accuracy as follows, and loss of data in the HDD 300 can be prevented.

  If abnormal writing due to particles being added under the head or accidental writing failure when the head could not completely write data to the sector, the access time temporarily exceeds the threshold, but again, By overwriting the same command issue number area, the access time is recovered, and thereafter the operation is performed with a normal access time. In such a case, it can be determined that the warning counter is reset and the HDD 300 has returned to normal operation.

  In addition, when the function is restored by the generation of the alternative sector, if it is single-shot, the function of the HDD 300 is normalized and the error count is not accumulated thereafter. It is considered that a write error has occurred due to a defective head. This is because the error counter is generated in the segment before and after the problem sector, or the error count is added when the access time exceeds the alternative sector occurrence prediction threshold and the occurrence of the alternative sector is considered. It can be determined that the failure of the HDD 300 is near.

  Further, a failure due to the life of the HDD 300 can be determined by checking whether or not the range in which the access time in which the warning counter is registered exceeds the threshold value is generated before and after the specific segment. In this case, it can be judged from the fact that the scatter diagram of FIG. 12B is expanding compared with the normal state.

  In particular, if a predetermined number (for example, two) or more of the warning counter occurs in the re-access of the command issue number area in which the warning counter is registered, recovery due to write failure is not expected, and there is a possibility of data writing abnormality in this command issue number area. Therefore, it is possible to more accurately determine that the failure is near by shifting from the warning counter to the error counter.

  In this way, the warning counter is accumulated and shifted to the error counter, and by analyzing the cause of the alternative sector of the HDD itself that can be detected from the SMART information at the warning counter occurrence position, in addition to the general failure determination of the HDD 300, time This makes it possible to more accurately determine obstacles that progress over time.

  The error counter is finally used to determine from the accumulated warning counter value that a threshold value for failure prediction has been exceeded and to notify the HDD 300 of a stop warning by an indicator lamp or the like. This may be a warning by a buzzer or the like, and may be linked with a stop function of the present system.

  Thus, by deriving a threshold value from the maximum access time during normal operation of the HDD 300, the range of access time during normal operation of the HDD 300 can be known, so abnormal operation can be accurately performed in the command issue number area where the access time exceeds the threshold value. The failure prediction of the HDD 300 can be performed with high accuracy. In this embodiment, processing is performed using the number of alternative sectors in the SMART information, but this can be said to be an index indicating the degree of failure of the recording medium. It can also be said that it is an index indicating the amount of resources used in the handling process related to the defect of the recording medium. The same processing can be performed using such an index, not limited to the number of alternative sectors.

  The HDD failure prediction apparatus 100 according to the second embodiment has the following characteristics.

  During operation, the HDD is accessed in command issue number units (specific capacity units) so that the access start position and transfer capacity are the same as those in the initial measurement, and the access time and SMART information changes corresponding to the command issue numbers ( In particular, failure prediction is performed based on the increase in alternative sectors.

  When the access time exceeds the threshold, the presence / absence of an increase in the alternative sector is determined, and when there is an increase, the warning notification is given a grace compared to when there is no increase. That is, when a replacement sector is newly assigned and the number of replacement sectors in the entire HDD is small, no warning is issued.

  When alternative sectors are concentrated in a narrow range of neighboring segments, a warning with a higher degree of seriousness is issued than when the alternative sectors are distributed over a wide range.

  When the access time corresponding to a certain command issue number exceeds a threshold, the warning counter corresponding to that command issue number is increased, and if the access time is less than the threshold after the second time, the warning counter is reset. Thereby, a temporary access delay and a permanent access delay can be distinguished.

  Access time measurement using an access pattern that resets the head position for each access time measurement of each area (command issue number) of the HDD so that the access conditions are the same during initial measurement and operation. I do.

  For the data obtained by measuring the access time of each area (command issue number) of the HDD, the local maximum value (local maximum value) of the access time is calculated for a plurality of adjacent areas, and abnormality detection is performed based on this value. Set the threshold.

(Embodiment 3)
In the HDD failure prediction apparatus 100 according to the third embodiment, the HDD SMART information is used as in the second embodiment, and a failure is predicted based on the access time based on a scatter diagram of access times according to a specific access pattern. A method of increasing the accuracy of failure prediction by determining a threshold value used at the time is adopted. In the third embodiment, if the failure is minor, a method of forcibly urging the generation of an alternative sector, which is a failure recovery function unique to the HDD, and recovering the access function of the HDD is adopted.

  The configuration and operation of the HDD failure prediction apparatus 100 according to the third embodiment are different from the warning determination processing in step S555 in the failure prediction operation in step S404 by the control unit 30 in that the recovery flag is set and the recovery processing is performed. Except for this, the configuration and operation of the HDD failure prediction apparatus 100 according to Embodiment 2 are the same. Here, the description of the configuration and operation common to the second embodiment will be omitted as appropriate, and the configuration and operation different from those of the second embodiment will be described.

  The control unit 30 measures the command execution time (access time) when data is written from the HDD controller 10 to the HDD 300, reads the threshold value stored in advance in the abnormal value DB storage unit, and sets the current access time to this threshold value. It is determined whether or not it exceeds.

  When it is confirmed that the access time exceeds the threshold, the control unit 30 stores the LBA, the access time and the SMART information at that time when the access time exceeds the threshold, and a warning (warning) counter and an error counter described later. The failure of the HDD 300 is predicted based on the transition of the warning counter and the error counter. When it is predicted that the life has been reached, an HDD stop warning is issued.

  Furthermore, if there is a possibility that the LBA whose access time exceeds the threshold will be recovered due to a failure factor, the control unit 30 notifies the host 200 of the LBA. Based on the information, the host 200 copies the data existing in the failed LBA to a normal area, and then sews the interval between normal file access operations to the failed LBA. Thus, unspecified data is written in accordance with the access pattern shown in FIG. 11 to force the transition to the alternative sector. At this time, the control unit 30 can confirm the position of the sector pending for substitution processing from the SMART information and the access time at that time, and can confirm that the sector has been shifted to the substitution sector. By forcibly shifting, the HDD can be returned to normal operation.

  The Current Pending Sector Count (hereinafter referred to as “alternative processing pending sector count”) in the SMART information is an important value for knowing the failure of the HDD as in the alternative sectors. In the actual operation of the HDD, a substitute sector does not occur suddenly. In many cases, a substitute processing pending sector is first generated in order to monitor the failure state. If it is confirmed that a failure of the same level as the previous occurrence occurs when the sector is accessed again, the HDD performs a replacement process and generates a replacement sector. However, as with the alternative sector, the number of sectors pending for substitution processing in SMART information simply indicates the number of occurrences, and the sector in which the substitution processing pending is generated in real time during HDD operation. There was no way to know. In the existing technology, the relationship between the alternative processing pending sector and the alternative sector is unknown, and it cannot be determined whether the alternative processing pending sector and the alternative sector occur in the same sector. For this reason, SMART information is mostly used for analysis of causes after an internal failure of the HDD occurs in existing technologies.

  If the write failure actually progresses, the failure does not occur and the replacement sector does not suddenly occur. If the HDD manufacturer has failed to write to the sector that was originally being written after the specified number of retries, the replacement process is performed first. If the sector is held as a pending sector and cannot be written at the time of accessing again, the process shifts to generation of a substitute sector. When the process shifts to the alternate processing pending sector, the HDD system does not have a failure in all the data of the sector to be written, and there is still room to determine that there is a possibility that the data can be written. It may be judged as a typical symptom. Actually, even when reading / writing is performed on a sector that is on hold of alternative processing, it often takes longer access time than a normal sector, but normal data may be read / written. However, in many cases, a failure will occur in the HDD in the future, and a transition will be made to an alternative sector. Therefore, if the occurrence of the alternative processing pending sector is accurately grasped, the failure of the HDD can be predicted more accurately.

  In addition, when the HDD system cannot read the sector information to be written at all, it shifts to the alternative sector without shifting to the alternative processing pending sector. It is possible to determine the level of the failure that has occurred by grasping whether the sector has been shifted to the alternative sector or whether the alternative sector has suddenly occurred.

  If it is possible to know in which sector of the HDD such an update of SMART information has occurred, it is possible to identify which failure has occurred in the HDD that is operating, and to determine which failure has occurred. It is possible to know at what fault level the sector is progressing with time, and it is possible to predict the arrival of a fault from the summation of fault distributions.

  Further, when the alternative processing pending sector is generated, although the access itself is possible, the HDD access exceeds the normal threshold value, so that the state is very unstable. However, since the access time returns to normal when the area shifts to the alternative sector, if the sector pending replacement processing can be transferred to the alternative sector at an early stage, the area can be used normally again.

  Therefore, after the data in the area where the alternative processing pending sector is generated is saved, the data is transferred to the alternative sector by forcibly writing to the sector where the alternative processing pending sector is generated. The unstable state can be solved and the HDD can be returned to the normal operation. This is called recovery processing.

  The recovery process is effective when the alternative processing pending sector occurs alone, but if multiple alternative processing pending sectors can be confirmed along with the alternative sector in a narrow area, the number of failures exceeding the number of alternative sectors can be confirmed. Therefore, it is determined as the HDD life.

  As described above, the HDD failure prediction apparatus 100 according to the third embodiment determines in which sector the SMART information changes when the access time exceeds the threshold with reference to the threshold of the access time during normal operation of the HDD. By grasping and tabulating, the progress of failures inside the HDD can be grasped more accurately. Since the SMART information and the state of the sector are correlated in real time to detect a sign of a failure occurrence, the failure can be predicted with high accuracy.

  Furthermore, in the HDD failure prediction apparatus 100 according to the third embodiment, which SMART information can be used in real time, which sector is a replacement processing pending sector and which replacement processing suspension sector has shifted to the replacement sector. Therefore, it is also possible to make effective use of the number of pending sectors for substitution processing in the SMART information, forcibly shift the substitution processing pending sectors whose access is unstable, and return to normal operation. Become.

  The outline of the access time of the HDD 300 is as described with reference to FIGS. 14A and 14B, but the processing time when the alternative processing pending sector is generated is added to the access time when the abnormality occurs. There is a relationship between the access time and the alternative processing pending sector and the change of the alternative sector. For the sector whose access time is increasing, the alternative processing pending sector or alternative sector is generated by the SMART information read immediately after this. Or the possibility of occurrence is high.

  The flowchart showing the failure prediction procedure performed by the HDD failure prediction apparatus 100 is the same as that shown in FIG. 7 described in the second embodiment. However, the failure prediction operation in step S404 includes a recovery operation. A recovery flag is set.

  In step S404, a failure prediction operation is performed when the HDD 300 is used. The control unit 30 monitors whether or not the access time of the command issue area exceeds the threshold value recorded in the abnormal value DB recording unit 40, and for the command issue number area exceeding the threshold value, the command issue number is stored in the abnormal value DB recording unit 40. And the access time and SMART information are recorded. In the method shown in FIG. 19, the current SMART information is compared with the SMART information when the access time exceeds the threshold value last time. Therefore, the abnormal value DB recording unit 40 temporarily stores the SMART information of the command issue number when the access time exceeds the threshold value last time, and uses it for comparison with the SMART information of the current command issue number.

  The control unit 30 counts the warning counter and the error counter recorded in the abnormal value DB recording unit 40. As a result, the sector in which the access time in the HDD exceeds the threshold and the distribution of changes in the SMART information in the sector are predicted to fail. Is reached, the HDD 300 is issued a stop warning and the process is terminated.

  In addition, when the failure is limited to the occurrence level of a single alternative processing pending sector due to the distribution of changes in SMART information, the control unit 30 saves data that uses the alternative processing pending sector once. Then, an arbitrary data write operation is performed in the alternative processing pending sector, and the alternative processing pending sector is forcibly shifted to the alternative sector, thereby stabilizing the subsequent operation of the HDD.

  The flowchart showing the detailed procedure of the failure prediction process in step S404 is the same as FIG. 8 described in the second embodiment, but the detailed procedure of the warning determination process in step S555 is different.

  In step S555, a warning determination process for detecting a non-fatal failure is executed. If the failure is such that the SMART information is not updated while the access time exceeds the threshold, or if data other than the number of alternative sectors is updated in the SMART information in step S543, it is determined that the failure is not fatal. Since there is something that detects the occurrence of some abnormality, warning processing is performed and the warning counter is totaled. In the warning determination process, a recovery flag is set to perform a recovery process for forcibly shifting to the alternative sector for a sector that may recover the HDD function at the place where the alternative process is pending. The warning determination process of the third embodiment will be described in detail with reference to the flowchart of FIG.

  FIG. 17 is a flowchart illustrating a detailed procedure of the warning determination process according to the third embodiment.

  In the warning determination process, the relationship between the generation position of the number of pending SMART information pending sectors and the generation position of the sector whose access time exceeds the threshold is checked in the HDD, and if the distribution of the generation positions overlaps, An error counter is added so that the determination of failure prediction in the determination process is accelerated more than usual.

  Also, by identifying the areas where an abnormality that cannot be detected by the change in the number of alternative sectors and the number of sectors pending for substitution processing is identified, and summing up the areas of the command issue numbers where the abnormality occurred, how the abnormal area is Judge whether it is distributed above.

  From these results, if the alternative processing pending sector is not continuously generated in the specific area, it is not concentrated in the distribution, and it is judged as an accidental failure, by setting a recovery flag, Prompts the recovery processing of the alternative processing pending sector. Although the recovery process will be described later, the recovery process is executed using the HDD access idle time independently of the main failure prediction process.

  The warning determination is performed by adding the value of the warning counter for each command issue number shown in FIG.

  The warning determination process is basically performed when the access time exceeds the threshold value as shown in step S535 of FIG. 8 and an access failure has occurred, but in step S543, the occurrence of an alternative sector cannot be confirmed in the SMART information. In addition, events other than the occurrence of a substitute sector with a high degree of urgency when a failure occurs are processed.

  In particular, the replacement of the SMART information with the alternative sector may be due to the fact that the alternative processing pending sector has shifted to the alternative sector, or due to the occurrence of a failure that suddenly causes the alternative sector. Even in this case, since it can be recognized that the HDD is seriously damaged, the error determination process is directly performed without performing the warning determination process, and the HDD is stopped as soon as possible.

  In the warning determination process, the distribution of alternative processing pending sectors is recorded, and by counting which alternative processing pending sectors have moved to the alternative sector, the progress of failures in the HDD is determined. For the occurrence of a single sector pending the alternative process where the spread of the distribution cannot be seen, a recovery flag is set to force the transition sector to the alternative sector and the HDD performance is restored. Prompt.

  In the life prediction, the replacement sector is not generated in the neighboring sector, but the replacement processing pending sector causing the access time delay exceeding the threshold gradually expands. The sector shifts to the alternative sector, and the alternative sector is rapidly expanded. Therefore, it is necessary to record the distribution of the access delay for each sector, and to determine how the sector exceeds the threshold in the scatter diagram band of FIG. is there.

  Therefore, in addition to the fact that the access time exceeds the threshold and no alternative sector has occurred, there is no transition from the alternative processing pending sector to the alternative sector in the neighboring sector, or the number of occurrences are totaled and the failure reaching the lifetime Predict progress.

  This is because, when an alternative processing pending sector occurs, the access time exceeds the threshold, but the HDD has not yet determined that it is a fatal error, and the HDD internal determination of whether or not to move to the alternative sector This is a state in which the threshold value is not reached. The occurrence of a specific number of such sectors in the neighboring area means that the sector in the vicinity of the area is not normal, and that many alternative processing pending sectors occur, there is a possibility that so many alternative sectors may occur. , Means that some kind of failure is progressing.

  However, since a sector that is pending for one substitution does not always move to the alternate sector at the next access, in order to see the state of access exceeding the first threshold, a warning counter is set in step S810. It is determined whether it is 0 (whether a warning counter already exists).

  If the warning counter is not 0 (NO in S810), the process proceeds to the same sector for the second time and thereafter, and the process proceeds to step S815. If the warning counter is 0 (YES in S810), the process proceeds to step S820.

  Thereafter, in steps S815 and S820, it is checked whether or not the number of alternative processing pending sectors has increased from the previous time.

  If the number of pending alternative processing sectors has not increased from the previous time in step S815 (NO in S815), there is a warning counter, and there is no new alternative pending sector, so only information on sectors that can be confirmed at this point in time. Since it is not possible to determine how the HDD failure has progressed, the process proceeds to step S825. Further, if the number of alternative processing pending sectors has increased from the previous time in step S820 (YES in S820), a warning counter does not exist and a new alternative processing pending sector has occurred, so this can be confirmed at this point. Since it is not possible to determine how the failure of the HDD has progressed only with the information regarding the sector, the process proceeds to step S825.

  In step S825, it is determined whether three or more warning counters have occurred in the first neighboring segment (a range of five segments before the access target segment and five segments after the access target segment). When step S825 is YES, in the scatter diagram of FIG. 12B, there is a high possibility that the access time exceeds the threshold and the convergence of the band is spreading. The predetermined number such as five segments before and five segments after is an example of the number of sectors in which an access delay occurs and is included in the adjacent segment area where abnormality can be determined. It may be used.

  When three or more warning counters (three warning counters with a value greater than 0) are confirmed in the first neighboring segment (YES in S825), the failure to shift to an alternative sector in the first neighboring segment is expanded. In step S835, the error counter is incremented by one.

  In step S845, the warning counter is 0 (YES in S810), and no alternative process pending sector has occurred (NO in S820). In step S825, the occurrence position of the alternative process pending sector and the previous This process is executed when the relevance of the occurrence position of the warning counter is not found (NO in S825). At this time, 1 is added to the warning counter of the sector currently being examined, and the warning determination process is terminated.

  In step S815, if the alternative process pending sector has increased from the previous time (YES in S815), the warning counter is not 0, so the second and subsequent access delays have occurred, and a new alternative process pending sector has also occurred. Therefore, it is considered that the failure has progressed, and the process proceeds to step S830.

  In step S830, it is confirmed whether or not 10 warning counters (predetermined number) have occurred in the first neighboring segment. In step S825, the relationship between the occurrence of the alternative processing pending sector and the access delay is not clear, so it is determined that the failure progresses slowly. However, in step S830, since the relationship between the occurrence of the substitution processing pending sector and the access delay is clear, it is determined that the occurrence of the access delay is definitely due to the substitution processing pending sector. Therefore, when it is confirmed that a predetermined number or more of alternative processing pending sectors are generated in the first neighboring segment (YES in S830), a failure occurs that may cause the alternative processing pending sector to shift to the alternative sector. In step S840, 10 is added to the error counter so as to issue an HDD stop warning in the next error determination process.

  On the other hand, if there is no occurrence of the pending sector for substitution processing that concentrates in such a specific area (NO in S830), it is considered that the sector is pending for substitution processing alone. However, in the HDD internal processing in the state where the alternative processing pending sector is generated, the time required for access increases as the processing for determining whether or not to shift to the alternative sector for each access increases. There is a possibility of exceeding. In addition, although the data storage situation is actually very unstable, if it is not determined that the internal processing of the HDD should move to the alternative sector, a retry occurs for each access, and the access time of the HDD exceeds the threshold and is unstable. It becomes a state. Therefore, when it is confirmed that such a stand-alone alternative process pending sector is generated, the write process to the alternative process pending sector is forcibly performed, and the alternative process pending sector is shifted to the alternative sector. Eliminate unstable conditions. In step S850, a recovery flag is set to perform the process. Details of the recovery process will be described later.

  A flowchart showing the detailed procedure of the error determination process in step S560 is the same as that in FIG. 10 described in the second embodiment, but some supplementary explanations will be made.

  As described above, in step S830 of the warning determination process, since the relationship between the occurrence of the substitution processing pending sector and the access delay is clear, the substitution processing pending sector to the substitution sector in the first neighboring segment is determined. If the transition is progressing rapidly, the error counter is incremented by 10. Therefore, in the error determination process, the process proceeds to step S730, and a HDD stop recommendation is issued.

  The number of adjacent second neighboring segment areas that are 10 before and 10 after the segment to be accessed and the number of error counters in the second neighboring segment area that are used in step S720 and 5 or more are as follows: It is a value for clarifying that the alternative processing pending sector has shifted to the alternative sector in an extremely narrow range, and changes depending on the difference in processing capacity of the HDD alternative processing pending sector to shift to the alternative sector. It is necessary to increase this value in an HDD with a low threshold value for judging whether to shift from the alternative processing pending sector to the alternative sector, and this value is required for an HDD having a high judgment threshold value for moving from the alternative processing pending sector to the alternative sector. Can be reduced.

  If generation of five or more alternative sectors is confirmed in step S710, the process proceeds to step S730 to issue an HDD stop recommendation to stop the use of the HDD, but the threshold that the number of alternative sectors generated is five or more is: It may be variable according to the number of alternative processing pending sectors. For example, when the number of alternative processing pending sectors is less than 10, the threshold of the number of alternative sectors generated is 10, and when the number of alternative processing pending sectors is 10 or more, the threshold of the number of alternative sectors generated is 5 It is good. This is because as the number of alternative processing pending sectors increases, the threshold for the number of alternative sectors generated is lowered to facilitate the HDD stop recommendation. Alternatively, a comprehensive index combining the number of alternative processing pending sectors and the number of alternative sectors may be calculated, and an HDD stop recommendation may be issued according to the total index. For example, the total index z = αx + βy may be used in the determination in step S710, where x is the number of alternative sectors and y is the number of sectors pending for substitution processing. Here, α and β are predetermined values larger than 0, and typically satisfy α> β.

  FIG. 18 is a flowchart showing a detailed procedure of the recovery process. The sector recovery process is performed by a process such as a task different from the failure prediction process. Basically, the failure prediction processing is performed as part of the main read / write processing of the host 200, but the sector recovery processing is performed by monitoring the recovery flag when the recovery flag is set in step S850 in FIG. Is done by the task. The sector recovery process is performed asynchronously with the failure prediction process while the recovery flag is set.

  The sector recovery process is performed in the free time of the main process of reading / writing from the host 200 and does not interfere with the original reading / writing process. Basically, the recovery process only writes to a sector designated by a specific capacity unit, similar to the read / write process of the host 200, and does not require a large processing time.

  In step S905, the SMART information for performing the recovery process is read again from the HDD, and the number of alternative processing pending sectors and the number of alternative sectors for the sector area to be forcibly written are read.

  When data is present in an area where recovery processing is to be performed, data saving processing is required. In step S910, it is confirmed whether or not there is data in an area where recovery processing is to be performed. If data exists in the area to be recovered (YES in S910), the host 200 performs data avoidance processing in step S915. This may be a copy to another area on the same HDD, or may be a save for other media.

  Thereafter, in step S920, the counter of the number of times of forced writing to be performed is reset, and in step S925, the problem sector is written in a specific capacity unit.

  In step S930, the SMART information after writing is read. In step S940, it is checked whether the number of sectors pending substitution processing of the read SMART information is zero. When the alternative processing pending sector shifts to the alternative sector, or when the alternative processing pending sector is merely a temporary abnormality and returns to the normal sector, the number of alternative processing pending sectors becomes zero.

  For example, if the number of alternative processing pending sectors in this area is 1, if the shift to the alternative sector is performed by writing, the number of alternative processing pending sectors in the SMART information after the write processing is reduced by 1 and the alternative sector is increased by 1. To do. When the sector returns to a normal sector when the alternative processing pending sector is re-accessed, the number of sectors pending alternative processing is reduced by 1, but the number of alternative sectors remains unchanged. Alternatively, if it is determined in the HDD internal processing that the error state is minor enough not to shift to the replacement processing, the number of sectors pending for replacement processing does not change and the number of replacement sectors does not change.

  When the alternative process pending sector shifts to the alternative sector or the alternative process pending sector returns to normal (YES in S940), the recovery flag is cleared in step S950, and the recovery process ends. The replacement sector generated at this time is used for determination in the error determination processing, and even if the replacement processing pending sector shifts to the replacement sector and returns to the normal access time, the replacement sector seems to increase as a whole. If there is, an HDD stop warning is displayed in step S730 of the error determination process to prompt the HDD to stop. Therefore, the transition to the alternative sector from the alternative processing pending sector also operates as one of the failure prediction processes.

  If the write process has not shifted from the alternative processing pending sector to the alternative sector (NO in S940), if the write counter has not yet reached 5 (NO in S945), 1 is added to the write counter in step S935. Returning to step S925, the writing process is performed again.

  If the write count reaches 5 in step S945 (YES in S945), the shift from the alternative process pending sector to the alternative sector is not observed even though the write process is performed 5 times, and the alternative process is pending. Since there is a possibility that a fatal failure has occurred in which sector transfer processing is not performed, the process proceeds to step S955, and an error counter is displayed so that HDD stop warning display processing is immediately performed in step S730 of error determination processing. Is set to 10 and the recovery process is terminated.

  The maximum value of the number of forced writes varies depending on the level for recognizing abnormality of the HDD. The alternate processing pending sector basically shifts to the alternate sector when a failure of the same level or more occurs at the next access after the occurrence. The recovery flag is set when an access delay occurs at least twice in the warning determination process of FIG. 17 in the area, and the occurrence of the alternate processing pending sector can be confirmed at the second time. It is unlikely that it will migrate normally by writing. For example, there may be cases where the alternative sector is not migrated while being judged in this way, for example, when the alternative sector has already been used up and there is no alternative sector to be migrated. It is dangerous and it is necessary to perform stop warning display processing immediately. Therefore, the number of forced writings may be reduced if it is desired to make the judgment of the degree of risk strict, and the number of forced writings may be increased if the judgment of the degree of risk is eased.

  As described above, according to the failure prediction procedure performed by the HDD failure prediction apparatus 100 according to the third embodiment, the access time is normalized by forcibly shifting the alternative processing pending sector to the alternative sector, thereby stabilizing the HDD. Can be made. Further, failure prediction of the HDD 300 can be performed with high accuracy, and loss of data in the HDD 300 can be prevented.

  Even if an alternative processing pending sector occurs, if the function is restored by moving to the alternative sector, if it is a single occurrence, the function of the HDD 300 will be normalized and the error count accumulation will not proceed thereafter. If the alternative processing pending sector and the alternative sector occur consecutively, it is considered that a write error due to a scratch or a head defect has occurred. It is determined that the failure of the HDD 300 is close by the occurrence of an error counter in the segment before and after the problem-occurring sector, or by adding an error count due to the occurrence of a substitution processing pending sector and a substitution sector. can do.

  Further, since the generation position of the alternate processing pending sector is known, the single alternate processing pending sector can be immediately forced to the alternate sector by writing, and the subsequent operation of the HDD can be stabilized.

  Furthermore, regarding failures due to the lifetime of the HDD 300, it is possible to confirm whether or not the alternative processing pending sector in which the warning counter is registered and the command issue number whose access time exceeds the threshold have spread around the specific segment. Judgment can be made. In this case, it can be judged from the fact that the scatter diagram of FIG. 12B is expanding compared with the normal state.

  In particular, when the relationship between the plurality of alternative processing pending sectors in the command issue number area where the warning counter is registered matches the distribution of command issue numbers whose access times exceed the threshold, many alternative sectors rapidly increase over time. Therefore, it is possible to more accurately determine that the failure is near by accelerating the progress from the warning counter to the error counter.

  In this way, the warning counter is accumulated and the process proceeds to the error counter, and by analyzing the sector pending the substitution process of the HDD itself that can be detected from the SMART information at the warning counter occurrence position and the cause of the substitution sector generation, In addition to the failure determination, it is possible to more accurately determine a failure that progresses over time.

  As described above, according to the failure prediction procedure performed by the HDD failure prediction apparatus 100 according to the third embodiment, if the failure occurrence level of the substitution processing pending sector is slight, the migration to the substitution sector is forcibly performed. Promptly, the stable operation of the HDD can be continued. Furthermore, by identifying the sector where the failure has occurred, it is possible to predict the life of the HDD with high accuracy and to avoid losing data due to the failure of the HDD.

  The present invention has been described based on the embodiments. The embodiments are exemplifications, and it will be understood by those skilled in the art that various modifications can be made to combinations of the respective constituent elements and processing processes, and such modifications are within the scope of the present invention. . In this embodiment, the processing is performed using the number of alternative processing pending sectors and the number of alternative sectors in the SMART information. However, these are indices indicating the degree of a recording medium failure or a sign of failure. I can say that. That is, the first index (the number of alternative processing pending sectors) that is relatively light and the second index (the number of alternative sectors) that is relatively heavy are used. Further, by writing to the recording medium, the first index (the number of alternative processing pending sectors) may return to a normal value. As long as the index has such characteristics, it is possible to perform the same processing using data other than the number of alternative processing pending sectors.

  In the above description, the failure prediction technique has been described by taking a hard disk as an example of the disk. However, the failure prediction technique of the present embodiment can be applied to any magnetic disk or optical disk. Further, the failure prediction technique of the present embodiment can be applied not only to a disk but also to a recording medium such as a memory card.

  In the above description, the change in the number of alternative sectors is detected from the SMART information by taking the hard disk as an example. However, when the failure prediction technique of the present embodiment is applied to a recording medium other than the hard disk, the SMART information is replaced. Thus, any state information for monitoring and analyzing the reliability of the recording medium may be used to detect an access area defect or a change in some index indicating a sign of failure.

  In the above description, the failure prediction is performed using the warning counter that may be reset and the error counter that is not reset. However, the failure depends on whether the warning counter exceeds a predetermined number using only the warning counter. May be determined.

  10 HDD controller, 20 temporary storage unit, 30 control unit, 40 abnormal value DB recording unit, 100 HDD failure prediction device, 200 host, 300 HDD.

Claims (12)

A status information recording unit for storing status information of the recording medium when a location to be accessed is accessed in a specific capacity unit according to a predetermined access pattern to the recording medium;
When the access target location is accessed in the specific capacity unit according to the access pattern, the status information of the recording medium is acquired, and the status information of the recording medium is associated with the access target location in the status information recording unit. A failure prediction apparatus comprising: a control unit to be registered.   The failure prediction apparatus according to claim 1, wherein the predetermined access pattern is a pattern for accessing the access target portion after returning a head for recording data on the recording medium to an initial position. .   The status information of the recording medium is an index indicating a degree of failure of the recording medium, or an index indicating an amount of resources used in processing corresponding to the defect of the recording medium. 2. The failure prediction apparatus according to 2.   4. The failure prediction apparatus according to claim 3, wherein the status information of the recording medium is the number of alternative sectors.   5. The failure prediction apparatus according to claim 1, wherein the specific capacity is determined according to the number of sectors in one track of the recording medium.   The control unit acquires state information of the recording medium when an access time when the access target location is accessed in the specific capacity unit according to the access pattern satisfies a first predetermined condition. The failure prediction apparatus according to any one of claims 1 to 5.   Prior to the measurement of the access time, the control unit measures in advance the access time corresponding to the access target location and the access target location in the vicinity thereof, and uses the plurality of access times measured in advance The threshold value corresponding to the access target location is calculated, and when the access time when the access target location is accessed in the specific capacity unit according to the access pattern exceeds the threshold, the first predetermined condition is The failure prediction apparatus according to claim 6, wherein it is determined that the condition is satisfied.   The control unit adds an error counter associated with the access target location when the newly acquired status information of the recording medium has changed compared to the previously acquired status information of the recording medium. The failure prediction apparatus according to claim 1, wherein an error determination process for detecting an abnormality based on the error counter is executed.   9. The failure prediction apparatus according to claim 8, wherein the control unit detects an abnormality when a change in state information of the recording medium satisfies a second predetermined condition at the access target location. .   The control unit has less influence on abnormality detection than the error counter, even when the newly acquired status information of the recording medium has not been updated as compared to the previously acquired status information of the recording medium. Adding a warning counter associated with the access target portion, and executing a warning determination process of adding the error counter when the warning counter in a neighboring area including the access target portion satisfies a predetermined condition The failure prediction apparatus according to claim 8 or 9. Storing state information of the recording medium in a state information recording unit when the access target portion is accessed in a specific capacity unit by a predetermined access pattern to the recording medium;
When the access target location is accessed in the specific capacity unit according to the access pattern, the status information of the recording medium is acquired, and the status information of the recording medium is associated with the access target location in the status information recording unit. A failure prediction method comprising: a step of registering. Storing state information of the recording medium in a state information recording unit when the access target portion is accessed in a specific capacity unit by a predetermined access pattern to the recording medium;
When the access target location is accessed in the specific capacity unit according to the access pattern, the status information of the recording medium is acquired, and the status information of the recording medium is associated with the access target location in the status information recording unit. A failure prediction program that causes a computer to execute the registering step.

Images