JP5505176B2 - Magnetic disk device, performance evaluation method and performance evaluation program - Google Patents

Description

The present invention relates to a performance evaluation technique for a magnetic disk device using a plurality of magnetic disks, for example, a magnetic disk device that evaluates performance based on write time and write speed affected by crosstalk, a performance evaluation method thereof, and a performance evaluation program About.

  In a magnetic disk device, a plurality of magnetic disks are arranged adjacent to each other, and a RAID device is configured with a plurality of magnetic disks, so that recording data is made redundant. In the case where a plurality of magnetic disks are provided, it is known that the performance of the magnetic disk device deteriorates due to the influence of crosstalk generated between the magnetic disks.

  Regarding this crosstalk, it is known to apply a signal that transitions in the same phase or opposite phase to a path that is affected by crosstalk, and to measure the propagation delay time of the signal that propagates to the path under the influence of crosstalk (for example, Patent Document 1).

  Regarding the detection of crosstalk generated between wirings, a test method is known in which a test pattern in which a malfunction occurs due to the influence of the crosstalk is extracted and a test pattern in which the influence of the crosstalk appears is generated. For example, Patent Document 2).

As for crosstalk detection, it is known to generate a crosstalk detection data pattern, execute a write command or read command using the data pattern, and determine and detect crosstalk from the execution result. (For example, patent document 3).

JP 2002-257903 A JP 2001-305191 A JP 2006-338565 A

  Incidentally, the performance of a magnetic disk device using a plurality of magnetic disks deteriorates due to crosstalk. This crosstalk affects the detection output of the shock sensor mounted on the magnetic disk device, and interrupts the writing of data in the same way as when external factors such as vibration and shock act on the magnetic disk device. . For this reason, the data writing time and the writing speed are affected by the crosstalk, and the crosstalk becomes a factor of the performance degradation of the magnetic disk device. In addition, if crosstalk occurs frequently, the reliability of data writing is lowered.

Therefore, in view of the above problems, the purpose of the magnetic disk device, the performance evaluation method, and the performance evaluation program of the present disclosure is to take advantage of the fact that the data write time and write speed change due to crosstalk, thereby improving the performance of the magnetic disk device. There is to evaluate.

  In order to achieve the above object, the magnetic disk device of the present disclosure includes a write control unit that interrupts data writing in accordance with a change caused by crosstalk during data writing. For this reason, if crosstalk occurs, the writing speed is reduced due to the interruption of data writing. Therefore, the crosstalk evaluation unit measures the writing speed of the data pattern and evaluates the presence or absence of the influence of the crosstalk from the writing speed. That is, the first data pattern with less noise is written in a specific area of the magnetic disk, and the writing speed or writing time of the first data pattern is measured. Next, a noisy second data pattern is written in the specific area of the magnetic disk, and the writing speed or writing time of the second data pattern is measured. Then, the presence or absence of crosstalk is evaluated by comparing the writing speed or writing time of the first and second data patterns.

  In order to achieve the above object, a performance evaluation method for a magnetic disk device according to the present disclosure includes a data write interruption process, a first measurement process for measuring a write speed, a second measurement process for measuring a write speed, and an evaluation process. . In the data write interruption process, the data write is interrupted in accordance with the change due to the crosstalk when the data is written. In the first measurement step, the first data pattern with less noise is written in a specific area of the magnetic disk, and the writing speed or writing time of the first data pattern is measured. In the second measuring step, a noisy second data pattern is written in the specific area of the magnetic disk, and the writing speed or writing time of the second data pattern is measured. In the evaluation step, the presence or absence of crosstalk is evaluated by comparing the writing speed or writing time of the first and second data patterns.

In order to achieve the above object, a performance evaluation program for a magnetic disk device according to the present disclosure includes a data write interruption function, a first measurement function for measuring a write speed, a second measurement function for measuring a write speed, and an evaluation function. . In the data writing interruption function, data writing is interrupted in accordance with a change caused by crosstalk when data is written. In the first measurement function, the first data pattern with less noise is written in a specific area of the magnetic disk, and the writing speed or writing time of the first data pattern is measured. In the second measurement function, a noisy second data pattern is written in the specific area of the magnetic disk, and the writing speed or writing time of the second data pattern is measured. The evaluation function evaluates the presence or absence of crosstalk by comparing the writing speed or writing time of the first and second data patterns.

  According to the magnetic disk device, the performance evaluation method, and the performance evaluation program of the present disclosure, any of the following effects can be obtained.

  (1) Since the writing speed or writing time of data to a specific area is measured and compared, the presence or absence of the influence of crosstalk can be grasped, and the accuracy of performance evaluation of the magnetic disk device can be improved.

  (2) Since a sensor that detects external factors such as vibration and impact can be used to detect changes due to crosstalk, the performance evaluation can be simplified.

  (3) If such performance evaluation is used, it contributes to high performance of the magnetic disk device.

Other objects, features, and advantages of the present invention will become clearer with reference to the accompanying drawings and each embodiment.

1 is a diagram showing an example of a magnetic disk device according to a first embodiment. It is a flowchart which shows an example of the process sequence of performance evaluation. It is a flowchart which shows an example of the process sequence of write-in operation. It is a figure which shows an example of the RAID apparatus which concerns on 2nd Embodiment. It is a figure which shows the function of a RAID apparatus. It is a figure which shows the structural example of the hardware of a RAID apparatus. It is a figure which shows the structural example of a disk. It is a figure which shows the data writing to a medium. It is a figure which shows an example of a data pattern. It is a flowchart which shows an example of the process sequence of performance evaluation. It is a figure which shows generation | occurrence | production of crosstalk. It is a flowchart which shows the modification of the process sequence of performance evaluation. It is a flowchart which shows an example of the process sequence of the performance evaluation which concerns on 3rd Embodiment. It is a figure which shows an example of the magnetic disc apparatus based on 4th Embodiment. It is a flowchart which shows an example of the process sequence of performance evaluation. It is a figure which shows an example of the RAID apparatus which concerns on other embodiment.

[First Embodiment]

  The first embodiment will be described with reference to FIG. FIG. 1 shows an example of a magnetic disk device according to the first embodiment.

  The magnetic disk device 2 is an example of the magnetic disk device of the present disclosure, and includes a single or a plurality of magnetic disks, and configures a RAID (Redundant Array of Inexpensive Disks) device or the like. As shown in FIG. 1, the magnetic disk device 2 includes a magnetic disk 4, a magnetic head 6, a change detection unit 8, a write control unit 10, and a crosstalk evaluation unit 12.

  The magnetic disk 4 is an example of a data recording medium, and is composed of a single disk or a plurality of disks. The magnetic disk 4 has an infinite number of fine magnetic poles on the medium surface, and has a structure for storing data in the fine magnetic poles. The magnetic disk 4 is rotated by a spindle motor, and data is stored in the set storage area by the magnetic head 6 described above.

  The magnetic head 6 is an example of a write head that writes data to the magnetic disk 4. The magnetic head 6 receives a write output from the write control unit 10, glides the surface of the magnetic disk 4, and causes a magnetic field to act on the fine magnetic poles of the magnetic disk 4. Write data.

  The change detection unit 8 is an example of an indirect means for detecting crosstalk, and detects changes caused by crosstalk. Changes caused by crosstalk include vibrations and shocks caused by crosstalk. The detection output of the change detection unit 8 is input to the write control unit 10.

  The write control unit 10 is an example of a write control unit that controls data writing to the magnetic disk 4 by the magnetic head 6, and interrupts data writing according to the detection result of the change detection unit 8. The interruption of the data writing affects the data writing speed. If the interruption of the writing operation occurs frequently, the data writing time becomes long even for the same data, and the writing speed and the average writing speed are lowered.

  The crosstalk evaluation unit 12 is an example of a functional unit that evaluates crosstalk generated on the magnetic disk 4, and compares the data writing speed with respect to a specific area of the magnetic disk 4 to evaluate the presence or absence of crosstalk. . In other words, the interruption of data writing due to crosstalk changes the data writing speed. Therefore, by comparing the writing speed, the presence or absence of the influence of crosstalk can be evaluated.

  The crosstalk evaluation unit 12 includes a data pattern generation unit 14, a writing speed measurement unit 16, and a comparison unit 18. The data pattern generation unit 14 generates a first data pattern 20 with little noise and a second data pattern 22 with much noise. The second data pattern 22 with a lot of noise is information for generating pseudo crosstalk. The writing speed measuring unit 16 writes the data pattern 20 in a specific area of the magnetic disk 4 and measures the writing speed of the data pattern 20. The writing speed measuring unit 16 writes the data pattern 22 in a specific area of the magnetic disk 4 and measures the writing speed of the data pattern 22. The comparison unit 18 compares the writing speeds of the data patterns 20 and 22 and evaluates the presence or absence of crosstalk.

  Next, with reference to FIG. 2 and FIG. 3, the performance evaluation of the magnetic disk device will be described. FIG. 2 shows an example of a performance evaluation processing procedure, and FIG. 3 shows an example of a data writing processing procedure.

This processing procedure is an example of a performance evaluation method and performance evaluation program for a magnetic disk device. In this processing procedure, as shown in FIG. 2, the first data pattern 20 with less noise is written in a specific area X (for example, FIG. 8) of the magnetic disk 4 and the writing speed V ₁ of the data pattern 20 is measured. (Step S11). The average writing speed As may be used as the writing speed V ₁ . The average write speed As is a value obtained by dividing a predetermined data length by a write time, and the write time includes a write interruption time due to crosstalk.

Writing a second data pattern 22 noisy in a specific area X of the magnetic disk 4, to measure the writing speed V ₂ of the data pattern 22 (step S12). It may be used an average writing speed Bs in writing speed V _2. The average write speed Bs is a value obtained by dividing a predetermined data length by the write time, and this write time includes a write interruption time due to crosstalk.

Then, the presence / absence of crosstalk is evaluated by comparing the writing speed V ₁ with the writing speed V ₂ (step S13).

  As shown in FIG. 3, in the data write processing procedure in this performance evaluation, the data patterns 20 and 22 are written by normal data writing (step S21). In this data writing, during the data writing, the detection output of the change detecting unit 8 is monitored to determine whether there is a change due to crosstalk (step S22). If there is a change due to crosstalk (YES in step S22), the data writing is interrupted (step S23), and the process returns to step S22 to monitor the detection output of the change detector 8.

  If there is no change due to crosstalk (NO in step S22), it is determined whether writing of all data is completed (step S24). If writing of all data has not been completed (NO in step S24), the process returns to step S21 to continue writing data. Then, when the writing of all data is completed (YES in step S24), this process ends.

  According to the above embodiment, the following effects can be obtained.

  (1) Since the writing speed of data to a specific area is measured and compared, the presence or absence of the influence of crosstalk can be evaluated, and the accuracy of performance evaluation of the magnetic disk device 2 can be improved.

  (2) Since a shock sensor that detects external factors such as vibration and shock is affected by crosstalk, this shock sensor can be used for change detection to simplify performance evaluation.

  (3) If such performance evaluation is used, it can be used for separating the magnetic disk 4 that generates crosstalk, eliminating or improving a low-performance product, and contributing to higher performance of the magnetic disk device 2.

[Second Embodiment]

  The second embodiment will be described with reference to FIG. FIG. 4 shows a RAID device according to the second embodiment.

  The RAID device 30 is an example of a magnetic disk device according to the present disclosure, and is a device that realizes data redundancy using a plurality of magnetic disks. As shown in FIG. 4, the RAID device 30 includes a controller enclosure (CE) 32 and a disk enclosure (DE) 34.

  The CE 32 is an example of a control unit of the DE 34 and is connected to a host computer (HOST) 42 (FIG. 6), and performs calculation and control such as disk control, comparison of writing speed or writing time, crosstalk evaluation, and the like. Run.

  The DE 34 is an example of a disk drive unit. The DE 34 includes a disk unit 36 and a plurality of port bypass circuits (PBC: Port Bypass Circuits) 38 and 40.

  The disk unit 36 includes a plurality of Disk00, 01, 02. Disks 00, 01, 02... N are examples of the magnetic disk 4 described above, and are recording media on which data can be written and read.

  PBC38 (# 0) and PBC40 (# 1) are examples of a disk switching circuit, and a loop including selection of a disk to be written or read from Disk00, 01, 02... N or a target Disk. Perform switching, etc.

  Next, the CE 32 will be described with reference to FIGS. FIG. 5 illustrates an example of a functional unit of the RAID device, and FIG. 6 illustrates an example of a hardware configuration of the RAID device. 5 and 6, the same parts as those in FIGS. 1 and 4 are denoted by the same reference numerals.

  As shown in FIG. 5, the CE 32 includes the write control unit 10 and the crosstalk evaluation unit 12 described above. The functions of the write control unit 10 and the crosstalk evaluation unit 12 are as described above, and a description thereof will be omitted. The write control unit 10 is connected to the magnetic head 6 and the shock sensor 80. Since the function of the magnetic head 6 is as described above, the description thereof is omitted.

  The shock sensor 80 is an example of the above-described change detection unit 8 (FIG. 1), and detects vibration and shock due to crosstalk when data is written. The shock sensor 80 is originally used for detecting external factors such as vibration and shock. However, when crosstalk occurs during data writing, the crosstalk appears in the detection output of the shock sensor 80. That is, the shock sensor 80 can be used as the change detection unit 8 described above. The detection output of the shock sensor 80 is input to the write control unit 10 of the CE 32. At the time of data writing, the writing control unit 10 interrupts data writing according to the detection output of the shock sensor 80, that is, if the detection level exceeds a threshold value.

  The CE 32 includes a CPU (Central Processing Unit) 44, a timer unit 46, and a memory unit 48, as shown in FIG. The CPU 44 executes the program in the memory unit 48, fetches data, generates a data pattern, refers to time information from the timer unit 46, measures the average writing speed of data, and checks whether there is crosstalk by comparing the average writing speed Perform processing such as evaluation.

  The timer unit 46 is an example of a time measuring unit, and acquires time information such as a data writing time from the timer unit 46.

  The memory unit 48 is an example of a program or data storage unit, and is configured by a recording medium. In the memory unit 48, an OS (Operating System) and a performance evaluation program of the present disclosure are stored, and a data storage area 50 is set. In the data storage area 50, a data pattern storage area 52, an As storage area 54, a Bs storage area 56, and a Cs storage area 58 are set. Each data pattern is stored in the data pattern storage area 52 in order to generate a data pattern with less noise and a plurality of different data patterns with much noise. The As storage area 54 stores the measured average writing speed As. The Bs storage area 56 stores the measured average writing speed Bs. The Cs storage area 58 stores the set average write speed Cs.

  Since the magnetic head 6 and the shock sensor 80 are as described above, description thereof is omitted.

  Next, with reference to Disks 00, 01, 02... N and data writing, reference is made to FIGS. FIG. 7 shows a configuration example of the disk, and FIG. 8 shows a write operation.

  Each Disk 00, 01, 02... N is provided with a single or a plurality of recording media (Media) 60 as shown in FIG. As described above, the Media 60 has innumerable fine magnetic poles on the Media surface, receives a magnetic field from the magnetic head 6 described above, and stores data by magnetization.

  As shown in FIG. 8, Media 60 is rotated at a constant speed by a spindle motor (not shown) (arrow N). The magnetic head 6 supported on the tip of the cantilever 62 is slid onto the medium surface of the media 60, and when the specific area X of the media 60 passes directly below the magnetic head 6, data writing (write) is executed. During the data writing, the above-described shock sensor 80 detects external factors such as vibrations and shocks. Even when crosstalk is received, a change due to the crosstalk appears in the detection output.

  The detection output of the shock sensor 80 is taken into the CE 32, and when the detection level exceeds a predetermined level (threshold), the data writing operation is interrupted. In this case, when the data writing in the specific area X is temporarily interrupted, the retry of data writing to the specific area X waits for rotation until the specific area X passes directly under the magnetic head 6 again after going through the media 60. . This rotation standby delays the writing time and becomes a factor of lowering the average writing speed. That is, since an output is generated in the shock sensor 80 due to the crosstalk, the crosstalk causes the average data writing speed to be delayed.

  Next, FIG. 9 is referred with respect to a data pattern. FIG. 9 shows an example of a data pattern.

  In this embodiment, three types of first, second, and third data patterns A, B, and C are used. The data pattern A is a data pattern that does not cause crosstalk (its influence is small), and is a pattern in which all values are set to zero as shown in A of FIG. All the values are set to zero (All = 0) because the parallel data bus is always constant. Therefore, this data pattern A constitutes reference data with little influence of crosstalk.

  The data pattern B is a data pattern that is likely to cause crosstalk (the influence thereof is large). As shown in FIG. 9B, the data “00” and “ FF "is repeated data. This is a data pattern as means for causing crosstalk. In the case of an 8 [bit] parallel bus, if such a data pattern B is used, the 8 [bit] parallel data bus can always be changed simultaneously with the same polarity.

  The data pattern C is a data pattern in which crosstalk is likely to occur (the influence is large). As shown in C of FIG. 9, the data “0000” and “ It is repetitive data with “FFFF”. Although this is different from the data pattern B, this is a data pattern as means for causing crosstalk. In the case of a 16 [bit] parallel bus, if such a data pattern C is used, the 16 [bit] parallel data bus can always be changed simultaneously with the same polarity.

  In this embodiment, the reference pattern that does not cause crosstalk and the pseudo pattern that causes crosstalk are organized in 8 [bits] and 16 [bits], but the present invention is not limited to this. 32 [bit], 64 [bit], etc., and the number of patterns is not limited to 3, but may be 4 or more.

  Next, the performance evaluation processing procedure will be described with reference to FIGS. FIG. 10 shows an example of a performance evaluation processing procedure, and FIG. 11 shows a case where crosstalk occurs in the Disk00.

  This processing procedure is an example of the performance evaluation method of the magnetic disk device or the performance evaluation program of the present disclosure, and Disk 00 (FIG. 4) is selected as the performance evaluation target in order to make the description concrete.

In this processing procedure, as shown in FIG. 10, the data pattern A with less noise generation is written to the specific area X of the Disk 00 (step S101). After this data writing, the average writing speed As for data writing is measured (step S102). The average writing speed As, for example, a value obtained by dividing the amount of data to write end at write time T _A from the write start data pattern A.

The data pattern B with much noise generation is written (written) in the specific area X of the Disk00 described above (step S103). After the data writing, the average writing speed Bs for data writing is measured (step S104). The average writing speed Bs is, for example, a value obtained by dividing the amount of data write time T _B from write start data pattern B to write end.

In addition, the data pattern C with much noise generation is written (written) in the specific area X of the Disk00 described above (step S105). After this data writing, the average writing speed Cs for data writing is measured (step S106). The average writing speed Cs, for example, a value obtained by dividing the amount of data to write end at write time T _C from the write start data pattern C.

  Then, the average writing speed As and the average writing speed Bs are compared (step S107). If As = Bs is not satisfied (NO in step S107), crosstalk occurs in the Disk 00 in the data pattern B (FIG. 11). ), Disconnecting Disk 00 is executed (step S108).

  If As = Bs, no crosstalk has occurred in the data pattern B, so the average write speed As and the average write speed Cs are compared (step S109).

  If As = Cs is not satisfied (NO in step S109), since crosstalk has occurred in the disk 00 with the data pattern C (FIG. 11), the disk 00 is disconnected (step S108).

  If As = Cs (YES in step S109), since no crosstalk has occurred in the Disk 00 even in the data pattern C, this process ends.

  In this processing procedure, for example, the average write speeds As, Bs, and Cs are exemplified by values obtained by calculating the data amount from the write start to the write end of the data patterns A, B, and C by one data write. However, an average writing speed of several times of data writing may be used.

  Next, FIG. 12 is referred to regarding the processing procedure of this performance evaluation. FIG. 12 shows a modification of the performance evaluation processing procedure.

  This processing procedure clearly shows the processing contents of the CPU 44 (FIG. 6), and a size-related determination process is used to compare the average writing speed.

In this processing procedure, as shown in FIG. 12, the CPU 44 instructs to write the data pattern A with less noise generation into the specific area X of the Disk 00 (step S201). Simultaneously with the start of data writing, the CPU 44 instructs the timer unit 46 to start timing (step S202). The CPU 44 monitors the end of writing of the data pattern A to the Disk 00 (step S203). Simultaneously with the end of the data writing, the CPU 44 ends the time measurement of the timer unit 46, fetches the time ( _TAS ) required for the writing from the timer unit 46, and stores it in the As storage area 54 of the memory unit 48 (step S204).

Next, the CPU 44 instructs to write the data pattern B with much noise generation in the specific area X of the Disk 00 (step S205). The CPU 44 instructs the timer unit 46 to start timing (step S206). The CPU 44 monitors the end of writing of the data pattern B to the Disk00 (step S207). Simultaneously with the end of the data writing, the CPU 44 ends the time counting of the timer unit 46, fetches the time (T _BS ) required for the writing from the timer unit 46, and stores it in the Bs storage area 56 of the memory unit 48 (step S208).

Next, the CPU 44 instructs to write the data pattern C with much noise generation in the specific area X of the Disk 00 (step S209). The CPU 44 instructs the timer unit 46 to start timing (step S210). The CPU 44 monitors the end of writing of the data pattern C to the Disk00 (step S211). Simultaneously with the end of the data writing, the CPU 44 ends the time counting of the timer unit 46, fetches the time (T _CS ) required for the writing from the timer unit 46, and stores it in the Cs storage area 58 of the memory unit 48 (step S212).

Next, the CPU 44 compares the time T _AS in the memory unit 48 with the time T _BS (step S213). If T _AS <T _BS (YES in step S213), since crosstalk has occurred in the data pattern B, the disk 00 is disconnected (step S214), and this process ends.

If T _AS <T _{BS is} not satisfied (NO in step S213), the CPU 44 compares the time T _{AS in the} memory unit 48 with the time T _CS (step S215). If T _AS <T _CS (YES in step S215), since crosstalk has occurred in the data pattern C, the disk 00 is disconnected (step S214), and this process ends.

Further, unless T _AS <T _CS (NO in step S215), since the crosstalk is not generated, disk00 maintains the connection, the process ends.

  About the said 2nd Embodiment, the characteristic matter, the advantage, and a modification are enumerated below.

  (1) Based on the characteristic that the crosstalk affects the writing speed at the time of data writing, a condition for generating a pseudo crosstalk in the write operation is given, and depending on the presence or absence of the crosstalk, the magnetic disk device 2 or The performance of the RAID device 30 can be evaluated.

  (2) In the above embodiment, when writing data to the disk, the average writing speed is measured as the performance when writing “All-0” of the first data pattern A that is less affected by crosstalk.

  (3) As the change detection unit 8 (FIG. 1) that indirectly detects crosstalk, a characteristic that the shock sensor (vibration sensor) 80 is easily affected by crosstalk is used. Therefore, the second data pattern B uses “00” vs. “FF” (in the case of an 8-bit parallel bus) repetition in which all parallel data bus portions simultaneously operate with the same polarity. At this time, the performance with the write data is measured, and the presence / absence of crosstalk is evaluated based on the ratio of the previous “All-0” data pattern to the “simultaneous, same polarity repeated crosstalk” pattern. For this evaluation, the average writing speed is used. If there is no crosstalk, each data pattern shows the same performance, but if there is crosstalk, a performance difference occurs between the two data patterns. The presence or absence of this performance difference is detected, and not the absolute value performance depending on the disk structure such as the measured LBA (position) but the performance difference due to the difference in the data pattern in the same measured LBA (position). Thereby, it is determined whether the abnormal state has crosstalk or the normal state without crosstalk. Then, the disk where the crosstalk has occurred is cut off to achieve normalization.

  (4) In order to confirm the influence from the adjacent Disk, for example, the same data pattern as the data pattern written in the previous Disk 00 is written in the Disk 01 adjacent to the Disk 00, and the performance difference due to the difference in the data pattern is confirmed. To do. The crosstalk from the adjacent disk is confirmed by the average writing speed, and the disk where the crosstalk is generated is separated from the loop. Thereby, a RAID device without crosstalk can be configured.

  (5) The performance difference confirmation due to the difference in data pattern may be performed when the RAID device is turned on and at regular intervals, and crosstalk from disks other than adjacent disks can also be detected. In this case, the average writing speed of data writing between the Disk 00 and, for example, the Disk N that is not adjacent to each other may be measured and compared.

  (6) “00” vs. “FF” simultaneously operating with the same polarity is the case of 8 [bit] parallel bus, but the parallel data bus has data 8, 16, 32, 64. Also good. When a repeated write data data pattern of “00” vs. “FF” valid for 8 [bit] data width is used for a bus width of 16 [bit], “FF00” is repeated and the data bus is always the same. Value and is not effective as a crosstalk pattern. Therefore, in the data bus width of all the disk drives, a data pattern whose value varies with the same polarity may be executed sequentially and checked for each bus width.

  (7) In addition, in the second embodiment, the same effects as those described in the first embodiment can be obtained.

[Third Embodiment]

  Next, FIG. 13 is referred about 3rd Embodiment. FIG. 13 shows an example of a processing procedure according to the third embodiment.

  This processing procedure is an example of the performance evaluation method or performance evaluation program of the magnetic disk device according to the present disclosure. In this embodiment, a RAID device similar to that in the second embodiment may be configured. In this embodiment, the influence of crosstalk between adjacent disks is evaluated, and Disk 00 and Disk 01 (FIG. 4) are used as performance evaluation targets.

  In this processing procedure, as shown in FIG. 13, the data pattern A with less noise generation is written (written) in the specific area X of the Disk 00 (step S301). After the data writing, the average writing speed As for data writing is measured (step S302).

  The data pattern B with much noise generation is written (written) in the Disk00 specific area X and Disk01 described above (step S303). After the data writing, the average writing speed Bs for data writing is measured (step S304).

  Further, the data pattern C with much noise generation is written (written) in the Disk00 specific area X and Disk01 described above (step S305). After this data writing, the average writing speed Cs for data writing is measured (step S306).

  Then, the average writing speed As and the average writing speed Bs are compared (step S307), and if As = Bs, since crosstalk has occurred in the Disk 00 in the data pattern B (FIG. 11), the Disc 00 is disconnected. Execute (Step S308).

  If As = Bs, no crosstalk has occurred in the data pattern B, so the average write speed As and the average write speed Cs are compared (step S309).

  If As = Cs is not satisfied (NO in step S309), since crosstalk has occurred in the disk 00 in the data pattern C (FIG. 11), the disk 00 is disconnected (step S308).

  If As = Cs (YES in step S309), since no crosstalk has occurred in the Disk00 even in the data pattern C, this process ends.

  Therefore, in the third embodiment, the same effects as those described in the first and second embodiments can be obtained.

[Fourth Embodiment]

  Next, a fourth embodiment will be described with reference to FIGS. 14 and 15. FIG. 14 shows an example of a magnetic disk device according to the fourth embodiment, and FIG. 15 shows an example of a performance evaluation processing procedure.

  In the first, second, and third embodiments, the writing speed and average writing speed of data writing are used for the performance evaluation (the presence or absence of crosstalk) of the magnetic disk device 2, whereas the fourth embodiment In the embodiment, the data writing time is used for crosstalk evaluation.

  As shown in FIG. 14, the magnetic disk device 2 of this embodiment includes a magnetic disk 4, a magnetic head 6, a change detection unit 8, a write control unit 10, and a crosstalk evaluation unit 12. In this case, the crosstalk evaluating unit 12 is an example of a functional unit that evaluates the crosstalk generated on the magnetic disk 4, and compares the data writing time with respect to a specific area of the magnetic disk 4 to evaluate the presence or absence of the crosstalk. To do. In other words, the interruption of data writing due to crosstalk changes the data writing time. Therefore, if the writing time is compared, the presence or absence of the influence of crosstalk can be evaluated.

  The crosstalk evaluation unit 12 of this embodiment includes a write time measurement unit 17 together with a data pattern generation unit 14 and a comparison unit 18. The writing time measuring unit 17 measures the writing time of the data patterns 20 and 22. Then, the comparison unit 18 compares the writing times of these data patterns 20 and 22 and evaluates the presence or absence of crosstalk.

  Since the other functional units are the same as those in the first embodiment, the description thereof is omitted.

  Next, FIG. 15 is referred for the performance evaluation of the magnetic disk device. This processing procedure is an example of a performance evaluation method and performance evaluation program for a magnetic disk device, and data write time is used for processing information.

In the processing procedure, the specific area X (for example, FIG. 8) of the magnetic disk 4 writes the first data pattern 20 with less noise, measuring the writing time T ₁ of the data pattern 20 (step S31). The average writing time may be used as the writing time T ₁ .

Writing a second data pattern 22 noisy in a specific area of the magnetic disk 4, to measure the writing time T ₂ of the data pattern 22 (step S32). It may be using the average write time for writing the time T _2.

Then, the presence / absence of crosstalk is evaluated by comparing the writing time T ₁ and the writing time T ₂ (step S33).

  Since the data write processing procedure in this performance evaluation is as shown in FIG. 3, its description is omitted.

  According to the above embodiment, since the time for writing data to the specific area is measured and compared, the presence or absence of the influence of crosstalk can be evaluated, and the accuracy of performance evaluation of the magnetic disk device 2 can be improved. As other effects, the same effects as those of the first and second embodiments can be obtained.

[Other Embodiments]

  (1) In the above embodiment, the presence / absence of crosstalk is determined using the crosstalk evaluation unit 12 in the magnetic disk device 2 or the CE 32 installed in the RAID device 30, but the present invention is not limited to this. . As shown in FIG. 16, the configuration may be such that the presence or absence of crosstalk is determined using a personal computer (PC) 82 connected to the RAID device 30. That is, the present invention is not limited to the computer inside the magnetic disk device 2, and the presence or absence of crosstalk is determined using an external processing device such as a host computer connected to the magnetic disk device 2, and the magnetic disk device 2 is evaluated. May be.

  (2) The RAID device 30 has been described as an example of the magnetic disk device 2 having a plurality of magnetic disks, but the magnetic disk device of the present disclosure may constitute a storage device other than the RAID device.

[Comparative Example]

  A RAID device configured with a disk mounted on a RAID device or the like, between adjacent disks, and a RAID device in which a module in the device and a disk drive are in a narrow space causes crosstalk, and the crosstalk causes a performance error in the RAID device. It is known that

  In a RAID device, the density of disk drives and the efficiency of access are improved. For example, LBA (Logical Block Addressing) is used for the access, and there is no need to be aware of the physical structure of the disk drive. In the disk drive, the transfer speed of the blocks arranged on the outer peripheral side of the disk becomes faster due to the influence of the fast peripheral speed, and the transfer speed of the blocks arranged on the inner peripheral side is affected by the slow peripheral speed, Tends to slow down.

  Access using the LBA can be easily executed without being aware of the physical position and control based on the physical position. However, since the execution is simple, the block arrangement is possible even in the same disk drive. The transfer speed and performance differ from place to place. The LBA control is a control that cannot recognize such a performance difference.

  In general, the disk is affected by the high peripheral speed in accessing the block on the outer peripheral side of the medium, and the transfer speed becomes high, so that the performance becomes high. On the other hand, access to the block on the inner peripheral side of the disk is affected by the peripheral speed that is slower than that on the outer peripheral side, and the transfer speed becomes lower, resulting in lower performance.

  In addition, the Disk stores data using fine magnetic poles on the medium surface. Since it takes a huge amount of money to manufacture a medium having no defect on the entire surface of the medium, it is accepted that the defect is less than a specific amount, and the defect part is distributed. For this reason, the defect portion is different in each disk, and even if it is a disk of the same model, its performance tends to be different. Therefore, it is difficult to detect the performance degradation by comparing the performance of the disk with a certain threshold value or comparing the disks.

  The magnetic disk device is equipped with a vibration (shock) sensor that detects vibration. Since this vibration sensor has a structure for detecting minute vibrations, it is affected by crosstalk. In RAID, the space between a plurality of disk drives is narrow, and adjacent disk drives are easily affected.

  It is assumed that magnetic disk devices of the same model are arranged adjacent to each other, and a circuit configuration that is not affected by crosstalk is assumed even in the case of adjacent arrangement. However, multi-vendor magnetic disk devices are being designed for stable distribution.

  When an external factor such as vibration or impact is applied to the magnetic disk device, there is a risk of shaking writing on adjacent tracks. Therefore, a vibration sensor is installed to detect external factors such as vibration and impact. The write operation is temporarily interrupted by the detection output of the vibration sensor to prevent malfunctions such as shaking on adjacent tracks.

  Regarding the performance degradation due to crosstalk generated in such a magnetic disk device, such a problem is solved by the magnetic disk device 2 and the RAID device 30 disclosed in the above embodiment, and further, a performance evaluation method and a performance evaluation program. .

  Next, the following additional notes will be disclosed with respect to the embodiment including the above-described examples. The present invention is not limited to the following supplementary notes.

(Appendix 1) A magnetic disk device including a magnetic disk,
A change detector that detects changes caused by crosstalk when writing data;
A write control unit for interrupting data writing according to the detection result of the change detection unit;
A first data pattern with less noise is written in a specific area of the magnetic disk, a writing speed or a writing time of the first data pattern is measured, and a second data pattern with a lot of noise is written in the specific area of the magnetic disk. A crosstalk evaluating unit that measures the writing speed or writing time of the second data pattern and compares the writing speed or writing time of the first and second data patterns to evaluate the presence or absence of crosstalk;
A magnetic disk device comprising:

(Supplementary Note 2) The crosstalk evaluation unit measures a writing / writing speed or a writing time of a third data pattern having a lot of noise different from the second data pattern in the specific area of the magnetic disk, 2. The magnetic disk apparatus according to appendix 1, wherein the presence or absence of crosstalk is evaluated by comparing the writing speed or writing time of the data pattern and the third data pattern.

(Supplementary Note 3) The magnetic disk includes at least a first magnetic disk and a second magnetic disk adjacent to the first magnetic disk,
The crosstalk evaluation unit writes a first data pattern with less noise to a specific area of the first magnetic disk and the second magnetic disk, and measures a writing speed or a writing time of the first data pattern, Writing a noisy second data pattern in specific areas of the first magnetic disk and the second magnetic disk, measuring a writing speed or a writing time of the second data pattern, and measuring the first and second 2. The magnetic disk apparatus according to appendix 1, wherein the presence or absence of crosstalk is evaluated by comparing the writing speed or writing time of data patterns.

(Supplementary Note 4) The crosstalk evaluation unit measures a writing / writing speed or a writing time of a third data pattern having a lot of noise different from the second data pattern in the specific area of the magnetic disk, and 4. The magnetic disk device according to appendix 3, wherein the presence / absence of crosstalk is evaluated by comparing the writing speed or writing time of the data pattern and the third data pattern.

(Additional remark 5) Furthermore, according to the evaluation result of the said crosstalk evaluation part, the magnetic disk control part which cuts off a magnetic disk,
The magnetic disk device according to appendix 1, appendix 2, appendix 3 or appendix 4, characterized by comprising:

(Appendix 6) A method for evaluating the performance of a magnetic disk device including a magnetic disk,
When writing data, a process of interrupting data writing according to changes due to crosstalk;
Writing a first data pattern with less noise in a specific area of a magnetic disk and measuring a writing speed or writing time of the first data pattern;
Writing a noisy second data pattern in the specific area of the magnetic disk and measuring a writing speed or a writing time of the second data pattern;
Comparing the writing speed or writing time of these first and second data patterns to evaluate the presence or absence of crosstalk;
A method for evaluating the performance of a magnetic disk drive, comprising:

(Supplementary note 7) Further, a step of measuring a write / write speed or a write time of a noisy third data pattern different from the second data pattern in the specific area of the magnetic disk;
Comparing the writing speed or writing time of the first data pattern and the third data pattern to evaluate the presence or absence of crosstalk;
The method for evaluating the performance of a magnetic disk device according to appendix 6, wherein:

(Supplementary Note 8) Further, the magnetic disk includes at least a first magnetic disk and a second magnetic disk adjacent to the first magnetic disk, and the first magnetic disk and the second magnetic disk. Writing a first data pattern with less noise in the specific area of the first area and measuring a writing speed or a writing time of the first data pattern;
Writing a noisy second data pattern in specific areas of the first magnetic disk and the second magnetic disk and measuring a writing speed or a writing time of the second data pattern;
Comparing the writing speed or writing time of these first and second data patterns to evaluate the presence or absence of crosstalk;
The method for evaluating the performance of a magnetic disk device according to appendix 6, wherein:

(Additional remark 9) Furthermore, the process of measuring the write-write speed or write time of the 3rd data pattern with much noise different from the said 2nd data pattern in the said specific area | region of the said magnetic disk,
Comparing the writing speed or writing time of the first data pattern and the third data pattern to evaluate the presence or absence of crosstalk;
The method for evaluating the performance of a magnetic disk device according to appendix 8, characterized by comprising:

(Supplementary Note 10) Further, according to the result of the evaluation, a step of separating the magnetic disk;
The method for evaluating the performance of a magnetic disk drive according to appendix 6, appendix 7, appendix 8 or appendix 9, wherein

(Supplementary Note 11) A performance evaluation program executed by a computer mounted on a magnetic disk device or a computer connected to the magnetic disk device,
When writing data, a function to interrupt data writing according to changes due to crosstalk,
A function of writing a first data pattern with less noise in a specific area of a magnetic disk and measuring a writing speed or a writing time of the first data pattern;
A function of writing a noisy second data pattern in the specific area of the magnetic disk and measuring a writing speed or a writing time of the second data pattern;
A function of comparing the writing speed or writing time of these first and second data patterns to evaluate the presence or absence of crosstalk;
Is realized by the computer. A performance evaluation program for a magnetic disk device.

(Additional remark 12) Further, a function of measuring the write speed or write time of a third data pattern having a lot of noise different from the second data pattern in the specific area of the magnetic disk;
A function of comparing the writing speed or writing time of the first data pattern and the third data pattern to evaluate the presence or absence of crosstalk;
The performance evaluation program for a magnetic disk device according to appendix 11, characterized by including:

(Supplementary Note 13) Further, the magnetic disk includes at least a first magnetic disk and a second magnetic disk adjacent to the first magnetic disk, and the first magnetic disk and the second magnetic disk. A function of writing a first data pattern with less noise in a specific area of the first area and measuring a writing speed or a writing time of the first data pattern;
A function of writing a noisy second data pattern in specific areas of the first magnetic disk and the second magnetic disk and measuring a writing speed or a writing time of the second data pattern;
A function of comparing the writing speed or writing time of these first and second data patterns to evaluate the presence or absence of crosstalk;
The performance evaluation program for a magnetic disk device according to appendix 11, characterized by including:

(Supplementary Note 14) Further, a function of measuring a write / write speed or a write time of a noisy third data pattern different from the second data pattern in the specific area of the magnetic disk;
A function of comparing the writing speed or writing time of the first data pattern and the third data pattern to evaluate the presence or absence of crosstalk;
14. The performance evaluation program for a magnetic disk device according to appendix 13, wherein

(Supplementary note 15) Further, according to the result of the evaluation, a function of separating the magnetic disk;
The performance evaluation program for a magnetic disk device according to appendix 11, appendix 12, appendix 13 or appendix 14, characterized by comprising:

As described above, the magnetic disk device of the present disclosure, its performance evaluation method, and the most preferable embodiment of the performance evaluation program have been described. However, the present invention is not limited to the above description, and It goes without saying that various modifications and changes can be made by those skilled in the art based on the gist of the invention described in the scope or disclosed in the embodiments for carrying out the invention. Needless to say, it is included in the range.

DESCRIPTION OF SYMBOLS 2 Magnetic disk apparatus 4 Magnetic disk 6 Magnetic head 8 Change detection part 10 Write control part 12 Crosstalk evaluation part 14 Data pattern production | generation part 16 Write speed measurement part 18 Comparison part 20, 22, A, B, C Data pattern 30 RAID apparatus 32 Controller enclosure 34 Disk enclosure 36 Disk unit 38, 40 Port bypass circuit

Claims (6)

A magnetic disk device comprising a magnetic disk,
A change detector that detects changes caused by crosstalk when writing data;
A write control unit for interrupting data writing according to the detection result of the change detection unit;
A first data pattern with less noise is written in a specific area of the magnetic disk, a writing speed or a writing time of the first data pattern is measured, and a second data pattern with a lot of noise is written in the specific area of the magnetic disk. A crosstalk evaluating unit that measures the writing speed or writing time of the second data pattern and compares the writing speed or writing time of the first and second data patterns to evaluate the presence or absence of crosstalk;
A magnetic disk device comprising:
The crosstalk evaluating unit measures a writing speed or a writing time of a third data pattern having a lot of noise different from the second data pattern in the specific area of the magnetic disk, and the first data pattern and the 2. The magnetic disk apparatus according to claim 1, wherein the presence or absence of crosstalk is evaluated by comparing the writing speed or writing time of the third data pattern.
The magnetic disk includes at least a first magnetic disk and a second magnetic disk adjacent to the first magnetic disk;
The crosstalk evaluation unit writes a first data pattern with less noise to a specific area of the first magnetic disk and the second magnetic disk, and measures a writing speed or a writing time of the first data pattern, Writing a noisy second data pattern in specific areas of the first magnetic disk and the second magnetic disk, measuring a writing speed or a writing time of the second data pattern, and measuring the first and second 2. The magnetic disk apparatus according to claim 1, wherein the presence or absence of crosstalk is evaluated by comparing the writing speed or writing time of data patterns.
The crosstalk evaluating unit measures a writing speed or a writing time of a third data pattern having a lot of noise different from the second data pattern in the specific area of the magnetic disk, and the first data pattern and the 4. The magnetic disk apparatus according to claim 3, wherein the presence or absence of crosstalk is evaluated by comparing the writing speed or writing time of the third data pattern.
A method for evaluating the performance of a magnetic disk device including a magnetic disk,
When writing data, a process of interrupting data writing according to changes due to crosstalk;
Writing a first data pattern with less noise in a specific area of a magnetic disk and measuring a writing speed or writing time of the first data pattern;
Writing a noisy second data pattern in the specific area of the magnetic disk and measuring a writing speed or a writing time of the second data pattern;
Comparing the writing speed or writing time of these first and second data patterns to evaluate the presence or absence of crosstalk;
A method for evaluating the performance of a magnetic disk drive, comprising:
A performance evaluation program executed by a computer mounted on a magnetic disk device or a computer connected to the magnetic disk device,
When writing data, a function to interrupt data writing according to changes due to crosstalk,
A function of writing a first data pattern with less noise in a specific area of a magnetic disk and measuring a writing speed or a writing time of the first data pattern;
A function of writing a noisy second data pattern in the specific area of the magnetic disk and measuring a writing speed or a writing time of the second data pattern;
A function of comparing the writing speed or writing time of these first and second data patterns to evaluate the presence or absence of crosstalk;
Is realized by the computer. A performance evaluation program for a magnetic disk device.

Images