JP2012038362A - Hard disk failure sign detection method - Google Patents

Abstract

In an information processing apparatus with a duplex system configuration, a hard disk failure sign is detected before an error occurs in the hard disk.
By referring to a history of execution time lengths of data writing to a hard disk 11 recorded in an execution time length recording area 211, a plurality of recent (for example, 10) are read (S31). Next, an average value of a plurality of read execution time lengths is calculated (S32). Next, it is determined whether or not the average value of the calculated execution time lengths of recent data is equal to or less than an allowable value (S33). If the average value exceeds the permissible value, an alarm notification transmission is instructed on the assumption that a failure sign has been detected (S34). For example, when the average value of the obtained recent execution time length is 12 seconds, an alarm notification transmission is instructed.
[Selection] Figure 4

Description

  The present invention relates to a hard disk failure sign detection method.

  In recent years, there has been a remarkable progress in increasing the capacity of hard disk drives (hereinafter referred to as HDDs or hard disks) mounted on information processing apparatuses. As a result, it is a welcome situation that a large amount of data can be stored in the HDD, but on the other hand, there is a concern that the stored data may be lost due to the failure of the HDD that is rarely experienced. In particular, the progress of increasing the capacity of HDDs means that lost data also has a large capacity.

  Therefore, it is necessary to replace the HDD before the data is destroyed. For this purpose, it is necessary to notify the maintenance person by some kind of alarm before the stored data is lost.

  In terms of alarming, some conventional HDDs also have a function called Self Monitoring Analysis and Reporting Technology (SMART). The SMART function stores HDD operation history, occurrence error history, etc. in the HDD, and compares these data with reference values to automatically issue warnings such as HDD replacement to higher systems such as personal computers and servers. To report. Since this SMART function is well known, further explanation will be omitted.

  Further, regarding the prior art, there is a technique disclosed in Patent Document 1 “Magnetic Disk Device Diagnosis Method”.

  According to this Patent Document 1, a disk device diagnosis method for the purpose of preventive maintenance of a magnetic disk medium, and when the number of occurrences of temporary failures is counted and a certain threshold value is exceeded during execution of disk medium diagnosis. A magnetic disk device diagnosis method for automatically reporting has been proposed.

  However, even in such a conventional technique, a hard disk failure cannot be foreseen until a temporary failure occurs in the hard disk, and unless there is a minor abnormality in the hard disk, there is no sign of a hard disk failure. There was a problem that I could not know.

JP 05-265663 A

  In the conventional HDD failure detection method described above, an HDD failure can be foreseen only when an error occurs in the HDD. Therefore, in the stage where no error including a minor error occurs, There is a problem that it is impossible to know in advance the signs that an abnormality or failure will occur.

  This problem is that an alarm is not notified until an error occurs in the HDD, so that the HDD replacement procedure cannot be performed before the error occurs. In particular, if an HDD failure occurs within a shorter time than the time required for HDD replacement after an error has occurred in the HDD, the data stored in the HDD will be lost, and the information processing apparatus will operate. Is a problem.

  An object of the present invention is to provide a hard disk failure sign detection method capable of detecting a hard disk failure sign before an error occurs in the hard disk in an information processing apparatus having a duplex system configuration.

  In order to solve the above problems, a hard disk failure sign detection method according to the present invention includes an information processing apparatus and one of the information processing apparatuses that performs information processing on behalf of the information processing apparatus when the information processing apparatus fails. Measuring the execution time length of writing or reading data to or from the hard disk of the other information processing device, the one information processing device storing the execution time length, The information processing apparatus reads the stored execution time length, and determines whether the execution time length exceeds a predetermined allowable value as a determination as to whether or not there is a sign of failure of the hard disk And a step.

  For example, the one information processing apparatus is an information processing apparatus that performs information processing when the other information processing apparatus fails, and stores the execution time length in the information processing apparatus that performs information processing when the failure occurs.

  According to the present invention, the execution time length of writing or reading data to the hard disk is measured and stored, the execution time length is read, and the determination as to whether or not there is a sign of a hard disk failure is the execution time length. Since it is determined whether or not a predetermined allowable value is exceeded, an information processing apparatus having a duplex system configuration can detect a sign of a hard disk failure before an error occurs in the hard disk.

  In addition, when the execution time length is stored in the information processing apparatus that performs information processing at the time of the failure, when the other information processing apparatus fails and it is necessary to refer to the execution time length of the information processing apparatus Even if it exists, the said execution time length memorize | stored in the information processing apparatus which processes information at the time of failure can be referred.

It is a block diagram of the information processing system which performs the hard disk failure sign detection method which concerns on this Embodiment. 10 is a flowchart of processing in which a standby system apparatus measures an execution time length of data writing to a hard disk in an active system apparatus. It is a flowchart of the process which calculates the allowable value of execution time length of data writing. FIG. 4A is a flowchart of processing for determining whether or not the execution time length of data writing exceeds the allowable value. FIG. 4B shows an alarm when the execution time length exceeds the allowable value. It is a flowchart of the process which transmits notification.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 is a configuration diagram of an information processing system that executes a hard disk failure sign detection method according to the present embodiment.

  The information processing system according to this embodiment includes two information processing apparatuses 1 and 2 that are connected to each other by a communication cable 3. Of these two information processing apparatuses having a redundant configuration, one is an active system apparatus and the other is a standby system apparatus and is operating. That is, the standby device performs information processing instead of the active device when the active device fails.

  The information processing apparatus 1 includes a hard disk 11 that records data, an execution time length measurement unit 12 that measures an execution time length of data writing or reading to the hard disk 21 of the information processing apparatus 2, and an indication of a failure of the hard disk 21. Whether or not there is a hard disk failure prediction detection unit 13 that determines whether or not the execution time length exceeds a predetermined allowable value, and an alarm notification when the execution time length exceeds the allowable value The alarm notification transmission part 14 which transmits. The hard disk 11 is provided with an execution time length recording area 111 for recording the execution time length.

  The information processing apparatus 2 also has the same configuration as the information processing apparatus 1, and an execution time for measuring the execution time length of data writing or reading to the hard disk 11 of the information processing apparatus 1 and the hard disk 11 of the information processing apparatus 1 A length measurement unit 22, a hard disk failure prediction detection unit 23 that determines whether or not the execution time length exceeds a predetermined allowable value as a determination as to whether or not there is a failure indication of the hard disk 11, and execution And an alarm notification transmission unit 24 that transmits an alarm notification when the time length exceeds the allowable value. The hard disk 21 is provided with an execution time length recording area 211 for recording the execution time length.

  The information processing apparatus 1 and the information processing apparatus 2 have the same configuration as described above, and are configured to perform the same operation when they are interchanged. However, in order to simplify the description, only processing during operation in which the information processing device 1 is an active device and the information processing device 2 is a standby device will be described below.

  Here, when the information processing device 2 measures the execution time length of data writing to the hard disk 11 of the information processing device 1 and detects a failure sign of the hard disk 21 from the measured execution time length, the alarm is issued. The method for transmitting the notification and the processing operation will be described sequentially with reference to the flowcharts of FIGS.

  First, the operation of a process in which the standby apparatus measures the execution time length of data writing to the hard disk in the active apparatus will be described with reference to FIG.

  The execution time length measurement unit 21 of the information processing apparatus 2 that is a standby system apparatus transmits a data write command signal and a data no-write signal simultaneously to the information processing apparatus 1 that is an active system apparatus (S11). ). The data write command is a command for executing a small data write of about several tens of megabytes (for example, 50 MBytes).

  The information processing apparatus 1 that has received these two signals writes data to the hard disk 11 depending on the data write command signal, and sends a response to the information processing apparatus 2 without doing anything depending on the no data write signal. To do.

  At this time, in the information processing apparatus 1, the response time after the data is written to the hard disk is later than the response time performed without doing anything. This is because the response transmission is delayed by the time for writing data to the hard disk. This time difference is the time required to write data to the hard disk, that is, the execution time length of writing.

  In this way, the two responses transmitted from the information processing device 1, that is, the response of the no-data-write signal and the response of the data-write command signal are received by the information-processing device 2 with the above-described time difference (S12). , S13).

  For example, if it takes 5 seconds to write 50 MBytes of data, the time difference between receiving the responses is 5 seconds.

  In the information processing apparatus 2 that sequentially receives the two responses, the execution time length measurement unit 22 calculates the time difference when the responses are received, that is, the execution time length of data writing to the hard disk 11 (in this example, 5 seconds) ( S14), recording in the execution time length recording area 211 (S15).

  Measurement of the execution time length of such data writing is performed periodically (for example, every 30 seconds), the execution time length of data writing to the hard disk is recorded, and the history is retained.

  As described above, here, only one of the measurement of the execution time length of data writing of the hard disk from the standby system device to the active system device has been described, but the hard disk from the active system device to the own system device (in the active system device) There is also a form in which the execution time length of writing to is measured. In this mode, the measurement of the execution time length of data writing is calculated by the difference between the data writing start time and the data writing completion time. Subsequent steps are the same as the steps for measuring the execution time length of the hard disk data write from the standby system device to the active system device described above.

  Next, the operation of the processing for calculating the allowable value of the data writing execution time length will be described with reference to FIG.

  A hard disk that is degraded takes a long time to write and read data, compared to a hard disk that is in good condition. As the time required for data writing increases, it is assumed that deterioration of the hard disk has progressed, and an allowable value for the execution time length of data writing is calculated as follows.

  The hard disk failure prediction detection unit 23 refers to the history of the execution time length of data writing to the hard disk 11 recorded in the execution time length recording area 211, and reads a plurality (for example, 10 pieces) from the beginning (S21). . Next, an average value of the read execution time lengths is calculated (S22). The average value is calculated in order to suppress variation due to various conditions when the execution time length is measured. Next, a value obtained by multiplying the calculated average value by a safety factor (for example, 2) is set as an allowable value for data writing time (S23). This means that the average value of the execution time length of data writing in the first 10 times is allowed up to twice the average value when the hard disk in good condition is used, for example. It is calculated as a value. For example, if the obtained average value is 5 seconds and the safety factor is 2, the allowable value is 10 (= 5 × 2) seconds.

  The permissible value calculated in this way, that is, the permissible value of the execution time length of data writing to the hard disk 11 is recorded in the hard disk 21 (S24).

  Next, referring to FIG. 4, (a) a process for determining whether or not the execution time length of data writing exceeds an allowable value, and (b) an alarm notification when the execution time length exceeds the allowable value. The operation of the transmission process will be described.

  As described above, the information processing apparatus 2 serving as the standby system apparatus regularly measures the execution time length of data writing to the hard disk 11 of the information processing apparatus 1 serving as the active system apparatus, and the recording is performed. Is stored in the execution time length recording area 211 as a history.

  As shown in FIG. 4A, the hard disk failure prediction detection unit 23 refers to the history of the execution time length of data writing to the hard disk 11 recorded in the execution time length recording area 211, and records a plurality of recent ( For example, 10 pieces are read out (S31). Next, an average value of a plurality of read execution time lengths is calculated (S32). The average value is calculated in order to suppress variations due to various conditions when measuring the execution time length of data writing.

  The execution time length measurement unit 22 determines whether or not the calculated average value of the execution time lengths of recent data writing is equal to or less than the allowable value calculated as described above (S33). Here, if it is equal to or less than the allowable value, the process is terminated assuming that the degree of deterioration of the hard disk is allowable. For example, when the allowable value is 10 seconds and the average value of the obtained recent execution time lengths is a value of 10 seconds or less, the process is terminated.

  However, if the average value exceeds the allowable value, the degree of deterioration of the hard disk is not allowed, that is, the warning notification transmission unit 24 is instructed to transmit the alarm notification on the assumption that a failure sign has been detected (S34). For example, when the average value of the recent execution time lengths obtained is 12 seconds, an alarm notification transmission is instructed.

  As shown in FIG. 4B, the alarm notification transmission unit 24 that has received the alarm notification transmission instruction from the execution time length measuring unit 22 (S41) transmits the alarm notification (S42), Notify deterioration.

  As described above, according to the present embodiment, the execution time length of data writing or reading to the hard disk is measured and stored, the execution time length is read, and whether or not there is a sign of a hard disk failure. As the determination, it is determined whether or not the execution time length exceeds a predetermined allowable value. Therefore, in the information processing apparatus having the duplex system configuration, a hard disk failure sign before an error occurs in the hard disk. Can be detected and the maintenance person can be informed.

  In addition, by storing the execution time length in the standby system device, even if a failure occurs in the active system device and it is necessary to refer to the execution time length for the active system device, the standby system device The stored execution time length can be referred to.

  The present invention is not limited to the present embodiment, and various modifications and changes can be made without departing from the spirit of the present invention. For example, the operation system apparatus may perform the failure sign detection method described above for the hard disk of the standby system apparatus. Further, data reading may be performed instead of data writing.

  The present invention records a program for realizing its function on a computer-readable recording medium in addition to that realized by dedicated hardware, and causes the computer to read the program recorded on this recording medium. , May be executed. The computer-readable recording medium refers to a recording medium such as a floppy (registered trademark) disk, a magneto-optical disk, and a CD-ROM, and a storage device such as a hard disk device built in the computer system. Furthermore, a computer-readable recording medium is a server that dynamically holds a program (transmission medium or transmission wave) for a short period of time, as in the case of transmitting a program via the Internet, and a server in that case. Some of them hold programs for a certain period of time, such as volatile memory inside computer systems.

1, 2 ... Information processing apparatus 11, 21 ... Hard disk 111, 211 ... Execution time length recording area 12, 22 ... Execution time length measurement unit 13, 23 ... Hard disk failure prediction detection unit 14, 24 ... Alarm notification transmission unit 3 ... Communication cable

Claims (5)

One of the information processing apparatus and the information processing apparatus that performs information processing on behalf of the information processing apparatus in the event of a failure of the information processing apparatus writes or reads data to or from the hard disk of the one information processing apparatus or the other information processing apparatus Measuring the execution time length of
The one information processing apparatus storing the execution time length;
Whether the one information processing apparatus reads the stored execution time length and whether the execution time length exceeds a predetermined allowable value as a determination as to whether or not there is a sign of failure of the hard disk A hard disk failure sign detection method comprising the steps of:   The one information processing apparatus is an information processing apparatus that performs information processing when the other information processing apparatus fails, and stores the execution time length in an information processing apparatus that performs information processing at the time of the failure. Item 5. A hard disk failure sign detection method according to Item 1.   3. The hard disk failure sign detection method according to claim 1, wherein the one information processing apparatus issues an alarm when the execution time length exceeds the allowable value.   A computer program for causing a computer to execute the hard disk failure sign detection method according to any one of claims 1 to 3.   A computer-readable recording medium on which the computer program according to claim 4 is recorded.

Images