JP2004146036A - Protective mechanism for magnetic disk, computer system provided therewith, method for protecting magnetic disk and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a protective mechanism for a magnetic disk which enhances the resistance of the magnetic disk to impacts by retreating a magnetic head according to the state of a magnetic disk unit and which meticulously sets the conditions for retreat of the magnetic head, and is effective and highly practicable. <P>SOLUTION: The protective mechanism is provided with a sensor 20 and sensor driver 31 which acquire information relating to a change in the environment for the magnetic disk unit 10, a shock manager 32 which analyzes the information acquired by the sensor driver 31 together with the history thereof and carries out impact prediction by judging the state in which the magnetic disk unit 10 is placed and an HDD filter driver 33 which controls the operation of the magnetic disk unit 10 including the retreat of the magnetic head on the basis of the result of the prediction by the shock manager. Further, the mechanism can also be provided with a diagnostic processing section for investigating whether the retreat of the magnetic head can be completed or not prior to the occurrence of the impact and for diagnosing whether there is a possibility of the occurrence of trouble in the magnetic disk or not. <P>COPYRIGHT: (C)2004,JPO

Description

The present invention relates to a protection mechanism for a magnetic disk drive, and more particularly, to a mechanism for protecting a magnetic disk from a shock or the like caused by dropping.

In portable information processing devices such as notebook computers, it is an important issue to protect the information processing devices from shocks such as accidental dropping while carrying or operating. In particular, a magnetic disk device generally used as a storage device of this type of information processing apparatus is structurally vulnerable to shock and vibration, and therefore it is desired to provide an effective protection means.

The magnetic disk drive reads and writes data by seeking the magnetic head on the rotating magnetic disk.If the magnetic head collides with the magnetic disk due to shock or vibration, the magnetic disk is damaged, and part or all of the data is restored. It may not be possible. Therefore, the impact resistance of the magnetic disk device can be improved by retracting the magnetic head from above the magnetic disk when an impact or vibration occurs.

As a protection mechanism of this kind in a conventional magnetic disk drive, a sensor for detecting that the magnetic disk drive is tilted, and a magnetic disk drive that tilts based on a detection signal of the sensor are provided in the magnetic disk drive. Evacuation control means for judging that the magnetic head is moved to the evacuation area, and detecting that the magnetic disk device is tilted as a sign of the fall of the magnetic disk device, and retreating the magnetic head from above the magnetic disk. (For example, see Patent Document 1).

保護 As a protection mechanism for the magnetic disk device, a magnetic disk device has been conventionally provided with a self-diagnosis function. A typical example of this self-diagnosis function is SMART (Self-Monitoring Analysis and Reporting Technology), which can obtain various information about the magnetic disk device, such as spin-up time, write error, seek rate, and temperature. (See, for example, Non-Patent Document 1).

Japanese Patent No. 2629548 (pages 1-3, FIG. 1) "Information Technology-AT Attachment with Packet Interface-5 (ATA / ATAPI-5)", [online], February 29, 2000, Technical Committee T13, [Search August 20, 2003], Internet <URL: http://www.t13.org/project/d1321r3.pdf> (p.39-41)

However, according to the conventional technique described in Patent Document 1, the degree of inclination of the portable terminal device on which the magnetic disk device is mounted is large or the portable terminal device on which the magnetic disk device is mounted falls due to the output of a pressure sensor that detects a change in weight of the magnetic disk device. Is detected, the magnetic head is retracted. However, the retracting condition is abstract and not clear. Therefore, as in the case of using a notebook computer on a lap, for example, it is impossible to deny the possibility that the magnetic head is unnecessarily evacuated by detecting the inclination of the portable terminal device that does not actually lead to the fall. In this case, each time the portable terminal device is tilted, the magnetic head is retracted from the magnetic disk, which may rather impair practical usability.
Further, Patent Document 1 does not describe the conditions under which the magnetic head is returned when the magnetic head is evacuated by mistake.

The self-diagnosis function of the magnetic disk device by SMART is not a function of determining whether a specific impact has damaged the magnetic disk device, but is a function of evaluating a failure as cumulative statistical information of the entire magnetic disk device. . Therefore, in the self-diagnosis of S.M.A.R.T., it is possible to acquire only information to the extent that the frequency of occurrence of an error exceeds a threshold value and a sign of a failure appears soon. Furthermore, even if a defective area occurs in the magnetic recording area on the magnetic disk due to the impact received by the magnetic disk drive, SMART will notice the existence of the defective area in normal use until the area is actually accessed. There is no.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an effective and practical protection mechanism for a magnetic disk drive, in which the evacuation conditions of a magnetic head are finely set.
Another object of the present invention is to realize a self-diagnosis for determining whether a magnetic disk device is damaged by an impact when the magnetic disk device receives an impact.

The present invention to achieve the above object is to provide an information acquisition unit for acquiring information on environmental fluctuations with respect to a magnetic disk device, and to analyze the information acquired by the information acquisition unit together with its history, and to set up the magnetic disk device. And a control unit for controlling the operation of the magnetic disk drive including the retraction of the magnetic head based on the prediction result by the shock prediction unit. It is realized as a protection mechanism for a magnetic disk.

Here, the state where the magnetic disk device is placed is specifically expressed as a tilt or swing of a housing of the magnetic disk device or an information processing device equipped with the magnetic disk device. These states are obtained from the acceleration generated in the magnetic disk device or the like detected by the acceleration sensor provided in the magnetic disk device or the like. The acceleration includes a linear acceleration and an angular acceleration, and the linear acceleration includes a static acceleration (gravitational acceleration) that fluctuates according to the attitude of the acceleration sensor and a force other than gravity received by a magnetic disk device or the like. Dynamic acceleration.

More preferably, the shock prediction means constituting the protection mechanism predicts that a shock will occur due to the change in the state when the placed state of the magnetic disk device changes in a predetermined pattern. In addition, it can be used for determining the state of the information processing apparatus in impact prediction, taking into account the history of input operations by a predetermined input device such as a keyboard and a mouse. On the other hand, if the change in the state falls within a certain range for a certain period, it is not predicted that an impact will occur due to the change in the state.
Further, after the magnetic head is retracted, the impact predicting means keeps the magnetic disk device in a stable state, that is, a stationary state or a posture fluctuation state within a certain range equivalent thereto continues for a certain period. When it is determined, the control means is notified of this. The control unit that has received the notification returns the retracted magnetic head. Here, the criterion (return condition) of the judgment by the shock prediction means is adaptively adjusted based on the information on the state of the information processing apparatus before the prediction of the occurrence of the shock, that is, the history of the information obtained from the information acquisition means. can do.

Further, when the magnetic head is retracted, the control unit does not execute a new access request to the magnetic disk device until the impact prediction unit determines that the magnetic disk device is in a stable state. Keep in queue. Thus, it is possible to prevent the magnetic head from returning while the magnetic disk device is in an unstable state, and to prevent a loss of an access request during that time.

According to another aspect of the present invention, there is provided a state determining means for determining a state of a magnetic disk device, and a control for controlling an operation of the magnetic disk device including retraction of a magnetic head based on a result of determination by the state determining means. The control means, when the state determination means determines that the magnetic disk device is in a state where there is a high possibility of receiving an excessive shock, issues an access request to the magnetic disk device once. The present invention is also realized as a magnetic disk protection mechanism characterized in that an access request having a small data size to be accessed is subdivided and transmitted to the magnetic disk device.

Further, another protection mechanism for a magnetic disk according to the present invention is such that when the control means determines that the magnetic disk device is in a state where there is a high possibility that the magnetic disk device is likely to receive an excessive impact, the magnetic disk device will The magnetic disk may be accessed by disabling the provided write cache function.
Furthermore, instead of disabling the write cache function, each time write data is generated in the cache memory, the control means writes the write data to the magnetic disk to empty the cache memory. You can also.

As described above, the magnetic disk protection mechanism according to the present invention not only predicts the occurrence of an impact and retracts the magnetic head but also, if an impact actually occurs, retracts the magnetic head before the occurrence of the impact. It is also possible to determine whether or not the magnetic head has been completed and diagnose whether or not there is a possibility that the magnetic head has collided with the magnetic disk and caused damage. The protection mechanism for a magnetic disk includes an impact prediction unit that predicts a shock that may occur in the magnetic disk device based on a change in environment with respect to the magnetic disk device, and a retracting unit that retracts the magnetic head based on a result of the prediction by the shock prediction device. Control means for controlling the operation of the magnetic disk drive including: when an impact actually occurs after the retraction of the magnetic head is started by the control means, whether the retraction of the magnetic head is completed before the occurrence of the impact Diagnostic means for making a determination of

More specifically, the diagnostic means includes a shock grace period, which is a time from the start of the magnetic head retreat operation to the time of the impact occurrence, and a retreat time required for the magnetic head retreat operation measured and held in advance. Judgment will be made by comparison. If the magnetic head has already been retracted when the retracting operation of the magnetic head is started under the control of the control means, the impact grace period is not compared with the retracting time, and the magnetic head is moved before the impact occurs. It is determined that the evacuation has been completed. Here, after the control unit issues a request command for controlling the magnetic disk device to perform the evacuation operation of the magnetic head, if the command completion notification is obtained within a predetermined time, the evacuation operation of the magnetic head is performed. At the start, it is determined that the magnetic head has already been retracted.
Further, the magnetic disk protection mechanism notifies a user of a diagnosis result by the diagnosis unit as to whether or not the magnetic head may have collided with the magnetic disk and caused damage, and performs recovery processing such as data backup. Can be encouraged.

The present invention that achieves the above object is also realized as a computer system having the above-described magnetic disk protection mechanism. Further, the present invention is also realized as a program that controls a computer and realizes the function of the above-described magnetic disk protection mechanism. This program can be provided by storing and distributing it on a magnetic disk, optical disk, semiconductor memory, or other recording medium, or distributing it via a network.

Further, the present invention that achieves the above-mentioned object, according to the present invention, uses a sensor to determine the state of the magnetic disk drive, and protects the magnetic disk by retracting the magnetic head according to the determination result. This is implemented as a method for protecting a magnetic disk. This magnetic disk protection method accumulates a history of information obtained from a sensor, analyzes the accumulated history and the last obtained sensor information to recognize a fluctuation pattern of the state of the magnetic disk device, When the occurrence of an impact on the magnetic disk device is predicted based on the content of the change in the state of the magnetic disk device, the magnetic head is retracted.

Further, another magnetic disk protection method according to the present invention includes a step of determining a state of the magnetic disk device being placed based on an output of a sensor, and a step of determining whether the magnetic disk device is likely to receive an excessive impact. When it is determined that there is, the operation control is performed such that the access request to the magnetic disk device is subdivided into an access request having a small data size to be accessed once and transmitted to the magnetic disk device. And retreating the magnetic head when the magnetic disk device is predicted to receive an excessive impact.

Further, instead of the above-described step of subdividing the access request, when it is determined that the magnetic disk device is in a state that is likely to receive an excessive impact, the write cache function of the magnetic disk device is invalidated. And controlling the operation so as to access the magnetic disk.
Further, instead of disabling the write cache function, every time write data is generated in the cache memory of the magnetic disk device, an operation of writing the write data to the magnetic disk to empty the cache memory is performed. May be.

The present invention is also realized as the following magnetic disk protection method. In this method for protecting a magnetic disk, first, a shock that may occur in the magnetic disk device is predicted based on a change in environment with respect to the magnetic disk device. Next, based on the result of the prediction, the operation of the magnetic disk drive including the retraction of the magnetic head is controlled. Then, when an impact actually occurs after the retraction of the magnetic head is started, an impact grace period, which is a time from the start of the retraction operation of the magnetic head to the time of the impact, and a magnetic head measured and held in advance By comparing the evacuation time required for the evacuation operation, it is determined whether or not the evacuation of the magnetic head is completed before the impact occurs.
With the above operation, it is possible to diagnose whether or not there is a possibility that the magnetic head has hit and damaged the magnetic disk.

According to the present invention configured as described above, the state of the magnetic disk device or the information processing device equipped with the magnetic disk device is monitored, and the retracting operation of the magnetic head is performed based on the finely set retracting conditions of the magnetic head. And an effective and highly practical protection mechanism for the magnetic disk drive.
Further, according to the present invention, when the magnetic disk device or the like is in an unstable state, the evacuation operation of the magnetic head can be quickly performed by preparing in advance.
Further, according to the present invention, when the magnetic disk device is actually shocked, a self-diagnosis for determining whether or not the magnetic disk device is damaged by the shock is realized, and the diagnosis result is provided to the user if necessary. The notification can prompt recovery processing such as data backup. As a result, if the magnetic disk was actually damaged, fragments of the coated magnetic surface of the magnetic disk that were scraped by the magnetic head would further damage the surface of the magnetic disk during subsequent operation of the magnetic disk device. Secondary failures can avoid damage that causes more data to be lost.

Hereinafter, the best mode for carrying out the present invention (hereinafter, an embodiment) will be described in detail with reference to the accompanying drawings.
First, the outline of the present invention will be described. INDUSTRIAL APPLICABILITY The present invention is directed to information processing apparatuses having a magnetic disk device as a storage device, particularly portable information processing apparatuses such as a notebook computer and a handheld computer, and these information processing apparatuses may fall. In this case, by detecting the sign and retracting the magnetic head from above the magnetic disk, the shock resistance performance of the magnetic disk device when the information processing device is actually dropped and a shock occurs is improved. In order to realize this, in the present invention, a state change such as occurrence of tilt or vibration of the information processing apparatus is monitored, and the state change is analyzed to detect a sign of a fall of the information processing apparatus. In addition, when a sign of a fall is detected, a preliminary process for quickly retracting the magnetic head is executed in the data transfer process.
Further, according to the present invention, when the information processing apparatus is actually shocked, based on information obtained by monitoring the state change or the like, it is determined whether or not the retracting operation of the magnetic head is in time, in other words, by the impact. A self-diagnosis is performed as to whether the magnetic disk device may be damaged.

FIG. 1 is a block diagram showing a configuration of an information processing apparatus to which a protection mechanism of a magnetic disk device according to the present embodiment is applied, and FIG. 2 is a diagram showing an appearance of the information processing apparatus.
As shown in FIG. 1, an information processing device 100 according to the present embodiment includes a magnetic disk device (HDD) 10 as a storage device, a sensor 20 for detecting the occurrence of tilt and vibration of the information processing device, A host computer (CPU and memory) 30 that controls the disk device 10 and the sensor 20 is provided. In the present embodiment, as shown in FIG. 2, a case where the information processing apparatus 100 is a notebook-type computer apparatus which is easy to carry and the magnetic disk apparatus 10 is mounted in a housing of the computer apparatus will be described as an example. . In this case, the host computer 30 is realized by a CPU and a main memory for data processing of the computer device itself. Further, since the magnetic disk device 10 is inclined together with the computer device or receives vibration or shock, the sensor 20 may be provided on the magnetic disk device 10 or provided on the housing of the computer device. good. Therefore, the following description will be made on the assumption that the information processing apparatus 100 is tilted or receives vibration or shock. However, depending on the apparatus configuration, the information processing apparatus 100 may be tilted or receive vibration or shock. It goes without saying that it applies.

Here, the situation where the protection mechanism of the magnetic disk device 10 according to the present embodiment is required will be described in detail.
FIG. 3 is a diagram schematically showing a general device configuration of the magnetic disk device 10. As shown in FIG.
Due to its structure, the magnetic disk device 10 moves the magnetic head 11 farther from the magnetic disk 12 than when the magnetic head 11 is on the magnetic disk 12 for reading and writing data (at the time of loading: position (a)). When it is at the predetermined evacuation position (at the time of unloading: position (b)), the resistance to shock and vibration is higher. Therefore, when the information processing apparatus 100 receives a large shock, the magnetic head 11 is retracted from the magnetic disk 12 in advance, whereby the shock resistance performance of the magnetic disk device 10 can be improved.
The most frequent example of the case where the information processing apparatus 100 receives a large impact is a case where the information processing apparatus 100 is dropped from a place where the information processing apparatus 100 is used, such as a desk or a knee. Therefore, it is conceivable that the magnetic head 11 is retracted from the magnetic disk 12 before the impact due to the fall is detected and the magnetic disk device 10 is protected before the information processing device 100 senses a sign of falling.

FIG. 4 is a diagram schematically illustrating a state change from a stable state to a fall of the information processing apparatus 100.
As shown in FIG. 4, the information processing apparatus 100 swings from a stable state placed on a desk or the like by being lifted by hand first (first stage). Next, as the initial operation of the fall, the swing of the information processing apparatus 100 becomes large and starts to rotate (second stage). Then, the information processing apparatus 100 falls freely (third stage), and finally collides with a floor or the like and receives a strong impact (fourth stage).
In the above process, the fall of the information processing apparatus 100 starts at the second stage. Therefore, it is quickly detected that the process has shifted from the first stage to the second stage, and the magnetic head 11 is retracted earlier. It is preferable to start the operation. Considering the case where the information processing apparatus 100 is operated on the knee, if this height is 50 cm (centimeter), the time required for the information processing apparatus 100 to fall from the height on the knee is about 320 msec (millisecond). ). Assuming that the initial operation of the second stage of drop is about 150 msec, the time required for the information processing apparatus 100 to fall from above the knee and generate an impact is about 470 msec.
In the magnetic disk device 10, the time required for the magnetic head 11 to retreat from above the magnetic disk 12 is about 300 msec in the case of a 2.5-inch HDD (hard disk drive). ), If the retracting operation of the magnetic head 11 is performed immediately, there will be sufficient time. However, in consideration of the operation control of the magnetic disk device 10 and the actual state of data transfer, it may not always be possible to immediately start the retreat operation of the magnetic head 11 after detecting the start of the fall.

Today, the mechanical structure of the magnetic disk drive 10 has been standardized to some extent, and it is difficult to add a special configuration for forcibly retracting the magnetic head 11 from above the magnetic disk 12 when a drop is detected. If a special design is made for that purpose, the production cost of the magnetic disk device 10 will increase. Therefore, in order to retreat the magnetic head 11, it is preferable to control the magnetic head 11 by software using an unload command as in the case of controlling the load / unload of the magnetic head 11.
However, since the HDC (hard disk controller) of the general magnetic disk device 10 operates in a single task, it cannot respond to the host computer 30 during operations such as data writing. Therefore, when the evacuation control of the magnetic head 11 is performed by the unload command, if the magnetic disk device 10 reads and writes data when the host computer 30 transmits the unload command to the HDC of the magnetic disk device 10, The execution of the unload command must be waited until the processing is completed, and the magnetic head 11 cannot be immediately retracted.

The magnetic disk device 10 generally includes a cache memory, and temporarily stores data to be written to the magnetic disk 12 in a high-speed cache memory (write cache) before writing to the low-speed magnetic disk 12. As a result, the response speed to the host computer 30 is improved. Normally, when the magnetic head 11 is unloaded, write data (Dirty Data) held in the cache memory is written to the magnetic disk 12 and the cache memory is flushed (emptied). Is unloaded. Therefore, when the host computer 30 transmits the unload command to the HDC of the magnetic disk device 10 and dirty data remains in the cache memory of the magnetic disk device 10, the execution of the unload command is performed until the cache is flushed. The user has to wait and cannot immediately retract the magnetic head 11.

From the above, it is assumed that the time required for the information processing apparatus 100 to fall from above the knee and generate an impact is about 470 msec, and the time required for retracting the magnetic head 11 from above the magnetic disk 12 is about 300 msec. However, there is not always enough time. Therefore, it is necessary to devise a method for starting the retracting operation of the magnetic head 11 immediately after the issuance of the unload command.

Further, in the protection mechanism of the magnetic disk device 10 according to the present embodiment, it is necessary to accurately determine the transition from the first stage to the second stage shown in FIG. It becomes.
When the information processing apparatus 100 is operated on a knee or the like, unlike the case where the information processing apparatus 100 is operated on a desk, the information processing apparatus 100 is likely to be subject to vibration or tilt. In this case, if the information processing apparatus 100 is simply tilted greatly or receives vibration of a certain strength or more as a condition for starting the retracting operation of the magnetic head 11, during operation of the information processing apparatus 100 on the knee, There is a possibility that the magnetic head 11 may be retracted by sensing the inclination or vibration of the information processing device 100 when the posture of the operator or the position of the information processing device 100 is changed. Naturally, if the magnetic head 11 is retracted, data cannot be read from or written to the magnetic disk device 10 and desired data processing cannot be temporarily performed. Further, when the information processing apparatus 100 is placed on the knee or the like and continuously tilts or receives vibration, the operation of the magnetic disk device 10 may not be restored, and the data processing may not be restarted.
Therefore, it is necessary to retreat the magnetic head 11 after analyzing the inclination and vibration of the information processing apparatus 100 and surely shifting the state of the information processing apparatus 100 to the second stage.

As described above, in the protection mechanism of the magnetic disk device 10, the magnetic head 11 is retracted by accurately determining that the state of the information processing device 100 has shifted from the first stage to the second stage shown in FIG. It is required to minimize the time from issuing the unload command for retracting the magnetic head 11 to starting the actual retraction operation.
Hereinafter, a system configuration and operation of the present embodiment that satisfies the above-described requirements will be described.

FIG. 5 is a diagram illustrating the system configuration of the protection mechanism of the magnetic disk device 10 according to the present embodiment.
Referring to FIG. 5, the present embodiment is obtained by a sensor 20 and a sensor driver 31 as information acquisition means for acquiring information on environmental changes such as tilt and vibration applied to information processing apparatus 100, and sensor 20. The shock manager 32 as analysis means and shock prediction means for analyzing the information obtained to determine the state of the magnetic disk device 10 and predicting the occurrence of a shock, and operating the magnetic disk device 10 based on the determination result of the shock manager 32. An HDD filter driver 33 is provided as control means for controlling.

The functions of the sensor driver 31, the shock manager 32 and the HDD filter driver 33 are realized by a program-controlled CPU in the host computer 30. Programs for controlling the CPU to realize these functions are provided by being stored and distributed on a magnetic disk, an optical disk, a semiconductor memory, or another storage medium, or distributed via a network. Then, after being stored in the magnetic disk device 10, it is read into the memory of the host computer 30 and executed by the CPU, thereby realizing the function of each component shown in FIG.

Further, the host computer 30 further controls the operations of the application 34 for performing various specific processes, the file system 35 provided by an operating system (OS), and the magnetic disk device 10 as normal functions. And an IDE device driver 37 that controls connection with the magnetic disk device 10. These normal functions in the host computer 30 are also realized by the program-controlled CPUs.

(4) Normally, when the application 34 accesses (reads and writes) a data file in the magnetic disk device 10, the access is performed via a file system 35 provided by the OS. The file system 35 manages how a data file composed of a group of data is actually arranged and stored in the magnetic disk device 10, and hides this from the application 34, so that the application 34 The use of the disk device 10 is simplified. The HDD driver 36 and the IDE device driver 37 actually access the magnetic disk device 10.

The HDD driver 36 that controls the operation of the magnetic disk device 10 is provided in accordance with an interface (IDE (Integrated Drive Electronics), SCSI (Small Computer System Interface), etc.) when the magnetic disk device 10 is connected to a host computer. Is connected to the magnetic disk drive 10 via a driver corresponding to the interface. In the present embodiment, it is assumed that the interface of the magnetic disk device 10 is IDE, and the IDE device driver 37 is used. The HDD driver 36 and the IDE device driver 37 access an I / O controller (for example, an IDE controller) to which the magnetic disk device 10 is connected under the control of the file system 35. Then, a predetermined I / O port of the I / O controller is operated so that the magnetic disk device 10 can transfer data at high speed in accordance with an instruction of the file system 35.

In the above configuration, the sensor driver 31 acquires information on the state of the information processing device 100 from the sensor 20. In the present embodiment, as the sensor 20, an acceleration sensor (accelerometer) that detects a change in acceleration applied on each of two axes (X / Y axis) or three axes (X / Y / Z axis) and performs analog output (Hereinafter, in this embodiment, a three-axis acceleration sensor is used.) Here, the acceleration sensor means an inertial sensor that measures a linear or angular acceleration. However, in general, an accelerometer often means a linear accelerometer. A group of angular accelerometers includes a gyroscope (angular velocimeter). Although the mounting method is slightly different, the present invention can be realized using any of the sensors.
The acceleration information at each time point constantly output from the sensor 20 is periodically (for example, every 10 msec) acquired by the sensor driver 31. The acceleration information acquired from the sensor 20 includes static acceleration (gravity acceleration) that varies according to the attitude of the sensor 20 (that is, the attitude of the housing of the magnetic disk device 10 or the information processing device 100 to which the sensor 20 is fixed). ) And dynamic acceleration that is generated by a force (such as an impact) externally applied to the housing of the magnetic disk device 10 or the information processing device 100. Generally, these combined values are output from the sensor 20.

The shock manager 32 analyzes the acceleration information acquired by the sensor driver 31 and the time of occurrence (time interval), and analyzes the inclination, movement, and rotation speed and acceleration of the housing of the information processing device 100 (or the magnetic disk device 10). Is calculated. These calculated values are stored in the shock manager 32 for a certain period of time as a fluctuation history up to immediately before.
Further, the shock manager 32 constantly monitors the current position of the housing of the information processing apparatus 100 (inclination, movement and rotation speeds and accelerations), and, based on this current state and its history, will It is determined whether or not an impact occurs (impact prediction).

Here, the function of the shock manager 32 will be described in detail.
FIG. 6 is a diagram showing a functional configuration of the shock manager 32.
As shown in FIG. 6, the shock manager 32 is a timer drive module that is periodically executed by a system timer, and includes a sensor monitor unit 61, an acceleration data history storage unit 62, a timer drive control unit 63, a keyboard / mouse event history. A storage unit 64 and an impact prediction unit 65 are provided.

Each time the sensor monitor unit 61 is periodically driven based on the system timer, the sensor monitor unit 61 acquires any or all of the acceleration levels of three axes (X / Y / Z axes) measured at that time from the sensor 20. The acquired information of the acceleration level (acceleration data) is used for impact prediction, but may be predicted for the acceleration level of the composite vector of each axis, or may be individually predicted for each axis. May be predicted with respect to the acceleration level of the axis. However, in the latter case, the prediction for determining the safety is made by the logical AND of the determination of each axis, and the prediction for determining the danger is made by the logical OR of the determination of each axis.
The acceleration data acquired by the sensor monitor 61 is sent to the shock prediction unit 65 and the acceleration data history storage unit 62.

The acceleration data history storage unit 62 stores the history of the acceleration sample data for a sufficiently long period of time, which is a major factor of a shock that can occur in a normal operation, as compared with the time elapsed in the fall. For example, if acceleration data corresponding to a history of 5 seconds at 100 Hz is sampled, the number of samples is 500 (= 100 × 5) samples on each axis.
The timer drive control unit 63 adjusts the actual drive cycle of the module that is periodically driven by a system timer or the like. For example, when the information processing apparatus 100 is in an unstable state that is likely to receive an excessive impact due to a drop or the like (high-risk mode described later), the information processing apparatus 100 is driven at 100 Hz and the information processing apparatus 100 is in a stable state. In the case (normal mode described later), control is performed such that monitoring and calculation are performed at 25 Hz on the assumption that monitoring is not required very frequently. As described above, by adaptively adjusting the actual driving cycle according to the state of the information processing apparatus 100, it is possible to reduce the load on the host computer 30 required for determining the state of the information processing apparatus 100 and detecting a drop. It is.
The keyboard / mouse event history storage unit 64 records keyboard and mouse events that can be acquired as normal system events for a certain period of time (for example, for the past 5 seconds like the history of acceleration data).

The status of the LCD panel, which is the display device of the information processing apparatus 100, indicates the open / close state of the LCD panel when the information processing apparatus 100 can open and close the LCD panel as in a notebook computer. In order to determine the use state of the device 100, it is directly obtained by the impact prediction unit 65. For example, when a change in the acceleration value is detected in the acceleration data acquired by the sensor monitor 61, if the LCD panel is open, the information processing apparatus 100 is used in an unstable state. I can judge. On the other hand, if the LCD panel is closed, it can be determined that the information processing apparatus 100 is being transported. In the latter case, there is a high possibility that the information processing apparatus 100 will fall erroneously. Therefore, the reference (threshold Th) of the impact prediction by the impact prediction unit 65 described below is lowered to increase Control such as retracting the magnetic head 11 in response to the sensitivity can be performed.

Similarly, an event due to a keyboard or mouse operation is stored in the keyboard / mouse event history storage unit 64 as described above, and is also directly acquired by the shock prediction unit 65, and is stored in the keyboard / mouse event history storage unit 64. It is used together with the history information to determine the use state of the information processing apparatus 100. For example, if an input operation using a keyboard or a mouse is performed before the information processing apparatus 100 falls, the user has fallen during use of the information processing apparatus 100, and the information processing apparatus 100 is unstable, such as on a knee. It can be determined that it was used in the state. On the other hand, if the input operation has not been performed for a certain period before the fall, it can be determined that the information processing apparatus 100 has fallen by an operation such as lifting the information processing apparatus 100 by hand. The results of these state determinations can be used as the evacuation conditions for the magnetic head 11 as appropriate. Further, if the input operation using the keyboard or the mouse by the user interaction is performed after the magnetic head 11 is retracted, the magnetic head 11 can be controlled to return under the condition.

The evacuation condition of the magnetic head 11 is determined by using the presence / absence of these input operations, the operation density (input amount per unit time), and the input pattern to adjust the impact prediction reference (threshold Th). It is also possible to further match the usage situation of the user in the above.
Specifically, when a predetermined input operation is performed by a keyboard or a mouse for a certain period of time from the current time to the past, the acceleration fluctuation detected at the current time is an unexpected fluctuation for a user who leads to the fall of the information processing apparatus 100. Rather, it is apparently a change during the operation, and even if it is a sign of a fall, it is considered that the user can quickly take measures to prevent it. Therefore, when such an input operation is performed, the threshold value Th is increased more than usual, the sensitivity to the change in the state of the information processing apparatus 100 is reduced, and the magnetic head 11 that is sensitive to the acceleration change accompanying the operation is increased. Control can be performed to suppress the evacuation operation. Further, by adding the operation density and the input pattern to the adjustment of the threshold value Th, for example, if an operation density equal to or higher than a certain value is not detected in a certain period going back from the present time to the past, the threshold value is obtained simply by detecting the input operation. It is also possible not to adjust Th.

In addition, in order to determine whether the user is intentionally operating the mouse, whether the trajectory of the mouse cursor is a monotonous linear trajectory, whether or not the movement speed of the cursor is higher than normal, and whether or not double-clicking is performed. It is also possible to selectively incorporate such evaluation into the adjustment of the threshold Th. When the user tries to prevent the information processing apparatus 100 from dropping and holds the information processing apparatus 100 at a dash, the user may unintentionally touch the mouse and perform mouse input. Performing selective use of mouse input is effective to prevent such unintended input from being mistaken for intentional input operation by a user during normal use. Similarly, regarding keyboard input, whether or not an appropriate type interval can be confirmed for a series of inputs can be used as a criterion for determining whether or not an input operation is intended by the user.

Each of the modules described above can communicate with the shock prediction unit 65 except for the timer drive control unit 63, and the shock prediction unit 65 sends a request to the HDD filter driver 33 to withdraw or return the magnetic head 11. And determination information for generating an operation mode switching notification (described later) and the like.

FIG. 7 is a diagram illustrating the internal function of the impact prediction unit 65 in more detail.
Referring to FIG. 7, the shock prediction unit 65 includes a immediately preceding fluctuation average calculation unit 71, an angle calculation unit 72, an angular velocity calculation unit 73, a fluctuation cycle detection unit 74, a free fall detection unit 75, a shock detection unit 76, and a retreat condition calculation unit. 77 and a return condition calculation unit 78.

The immediately preceding fluctuation average calculation unit 71 mainly uses the information from the acceleration data history storage unit 62 to calculate the average value of the acceleration (absolute value) obtained in the immediately preceding predetermined period or the integral value of the acceleration per unit time ( Speed). In addition, a predetermined statistic based on these may be calculated based on the average values of the acceleration and the speed. As the statistic, for example, a weighted average weighting the current side on the time axis, an average excluding a period in which data having a value outside a predetermined range is excluded, and a standard deviation of a fluctuation amount in the period are calculated. Can be.
Based on the calculated value, it is determined whether the information processing apparatus 100 is in a stable use state or is in an unstable state in which the information processing apparatus 100 is used while standing on a knee or standing. This makes it possible to adaptively change the sensitivity at the time of performing a shock prediction for a newly generated acceleration between a stable state and an unstable state. That is, if the average of the fluctuation of the acceleration and the speed (that is, the fluctuation of the state of the information processing apparatus 100) is within a certain range for a predetermined period, a certain degree of vibration, such as on a knee or in a moving car, may occur. It can be determined that this is not a sign of a fall, and control such as lowering the sensitivity for impact prediction can be performed.

算出 Furthermore, the calculated value can be similarly used for adapting the reference of the return condition when returning the magnetic head 11. That is, after the impact is predicted and the magnetic head 11 is retracted, no excessive impact is actually detected, and the magnetic disk device 10 is in a stable state (stationary or within a certain range equivalent thereto). Is determined to have continued for a certain period of time), a request to return the magnetic head 11 is made to the HDD filter driver 33, and the magnetic head 11 is returned. The return condition) is adjusted adaptively based on the state before the impact is predicted to occur, that is, the history of the inclination change and the dynamic acceleration change for a certain period obtained from the output of the sensor 20. For example, if the fluctuation width of the magnetic head 11 for a certain period before retreat is relatively large, it is considered that the information processing apparatus 100 was originally used in an environment with such a posture fluctuation (e.g., on the knee). Adjustment is performed such that the magnetic head 11 is returned under a relatively gentle return condition (even in a state where there is a slight change in posture).

The angle calculation unit 72 tracks the tilt displacement of the information processing device 100. More specifically, the current inclination is obtained by comparing the reference value with the current gravitational acceleration on the basis of a static gravity acceleration in a horizontal state which is obtained in advance. When the inclination is an angle equal to or greater than a certain angle, the angle does not become such an angle during use of the information processing apparatus 100, for example, immediately after the start of the fall (immediately after shifting to the second stage in FIG. 4). State). Therefore, when such an angular displacement has occurred, it is determined that an impact will occur according to a combination with a calculation result of the angular velocity calculation unit 73 described later. The acceleration value obtained from the sensor 20 is often a combination of a dynamic acceleration and a static acceleration. Therefore, it is necessary to extract a static acceleration by excluding a change in the frequency axis other than the low frequency band through a low-pass filter or the like.

The angular velocity calculator 73 is used for impact prediction in combination with the angular displacement calculated by the angle calculator 72 as described above. That is, it is determined that an impact is generated not only when the information processing apparatus 100 is tilted at a certain angle or more but also when the angular displacement occurs at a predetermined angular velocity. It should be noted that the occurrence of an impact is not necessarily predicted without accompanying a certain degree of angular displacement, but the allowable limit of the angular velocity according to the angle is set adaptively, and even when the inclination (angular displacement) is small, it is relatively small. When a large angular velocity is detected, it is possible to perform control such that it is determined that an impact occurs.
In addition, as a method of easily obtaining an angular velocity using a linear accelerometer, there is a method of sequentially calculating an angular displacement amount obtained from the linear accelerometer for each unit time, and substituting the obtained value as an average angular velocity.

The fluctuation period detection unit 74 monitors a specific acceleration or a fluctuation pattern of a numerical value (for example, speed) calculated based on the acceleration. Accordingly, even when there is an acceleration fluctuation of a level that cannot be predicted to generate an impact by itself, it is possible to predict that an impact will occur if the acceleration fluctuation occurs in a specific pattern. As a specific case, an acceleration fluctuation pattern detected when the user of the information processing device 100 is operating while walking the information processing device 100, or a case where the user slides down from the knee without extreme rotational movement The characteristic acceleration fluctuation pattern is a detection target.
When the gravitational acceleration suddenly changes suddenly and discontinuously, the free-fall detection unit 75 detects this and determines that the information processing apparatus 100 has started free-fall. If it is determined that a free fall of the information processing device 100 has occurred, it is determined that an impact will occur based on this.

(4) When the information processing apparatus 100 actually falls and suffers an excessive impact, the impact detection unit 76 determines this from the acceleration data. Then, the return condition calculation unit 78 should be judged on the return condition after actually receiving an impact, or should return promptly because the evacuation determination of the magnetic head 11 by the evacuation condition calculation unit 77 is incorrect. Provide information to determine

The evacuation condition calculation unit 77 receives information calculated by each module other than the above-described impact detection unit 76, comprehensively determines the information, and determines whether the information processing apparatus 100 is in an unstable state (the first stage in FIG. 4). ) Or whether the magnetic head 11 needs to be evacuated immediately after the fall is started (the second stage in FIG. 4). Then, in the latter case, a request to retract the magnetic head 11 is made to the HDD filter driver 33. In the former case, the IDE device driver 37 is notified of the switching of the operation mode for shifting to the high risk mode described later.
After the magnetic head 11 is evacuated by the notification from the evacuation condition calculation unit 77, the return condition calculation unit 78 inputs information calculated by each of the above-described modules including the shock detection unit 76, and comprehensively determines these. Then, it is determined whether or not the magnetic head 11 is to be returned. When returning the magnetic head 11, a request to return the magnetic head 11 is made to the HDD filter driver 33. Further, in accordance with the result of the determination, the IDE device driver 37 is notified of the switching of the operation mode for making a transition to the later-described normal mode.

FIG. 8 is a diagram illustrating an example of an operation algorithm by the evacuation condition calculation unit 77.
Here, only the angular velocity generated in the information processing apparatus 100 is simplified as the retracting condition of the magnetic head 11. The angular velocity at a predetermined point in time detected by the sensor 20 (that is, occurring in the information processing apparatus 100) is ω, the threshold for determining whether or not to request the evacuation of the magnetic head 11 is Th, Assuming that the integral value of the acceleration history accumulation period is v, the proportional constant that changes according to the current inclination (angular displacement) A is k _n , and the allowable minute fluctuation amount at a certain level is b, the threshold value Th is calculated by the following formula, for example. Is calculated.

Th = k _n * v + b
(However, k _n = 2 (| A | <π / 6), k _n = 1 (| A |> π / 6))

Referring to FIG. 8, when the angular velocity ω is greater than the threshold Th, the evacuation condition calculation unit 77 predicts that a drop will start (second stage in FIG. 4) and that an impact will occur, and the evacuation request of the magnetic head 11 is requested. For the HDD filter driver 33. Then, the magnetic head 11 is prevented from returning to the state before the shock is generated or while the shock can be continuously generated, and the magnetic head 11 is kept in the retracted state for a certain period of time. Start the evacuation timer (this is implemented by any suitable means such as counting a system timer).
When the angular velocity ω is larger than 閾 値 of the threshold Th, it is determined that the information processing apparatus 100 is in an unstable state (the first stage in FIG. 4), and the transition to the high risk mode described later is performed by the IDE. Notify the device driver 37. Then, a high-risk mode timer is started to maintain the operation state of the high-risk mode for a certain period of time.

FIG. 9 is a diagram illustrating an example of an operation algorithm performed by the return condition calculation unit 78.
Referring to FIG. 9, after the evacuation timer times out (elapse of a preset time), if the angular velocity ω is smaller than １ ／ of the threshold Th, the return condition calculation unit 78 determines that no impact occurs. Then, a request to return the magnetic head 11 is made to the HDD filter driver 33.
If the angular velocity ω is smaller than １ ／ of the threshold Th after the high-risk mode timer times out, it is determined that the information processing apparatus 100 is in a stable state, and the transition to the normal risk mode described later is performed by the IDE. Notify the device driver 37.

In the above operation algorithm, values such as １ ／ and ４ of the threshold value Th are merely examples, and other appropriate values may be used. Further, as described above, these values can be adaptively changed according to the calculation results of the immediately preceding fluctuation average calculation unit 71 and the fluctuation cycle detection unit 74. Further, in the actual shock prediction, not only the threshold value Th for the angular velocity ω but also the threshold value for the result of comprehensively judging information calculated by each module from the immediately preceding fluctuation average calculation unit 71 to the free fall detection unit 75 is used. Needless to say,
Further, the configuration and the operation algorithm of the shock manager 32 described above are merely preferred examples in the present embodiment, and the state of the information processing apparatus 100 is determined based on the output of the sensor 20 and the retracting of the magnetic head 11 is performed. Needless to say, various mounting modes can be adopted depending on the type of the sensor 20 and the like, as long as it controls the switching of the operation mode and the switching of the operation mode.

The HDD filter driver 33 is provided between the HDD driver 36 and the IDE device driver 37, and retreats the magnetic head 11 based on an instruction (retraction request and return request for the magnetic head 11) from the shock manager 32. Issue an unload command or a load command to restore. In addition, by controlling access to the magnetic disk device 10 by the HDD driver 36 and the IDE device driver 37, preparations are made to quickly retract the magnetic head 11 when the information processing device 100 falls. Specifically, regarding the access to the magnetic disk device 10, the HDD filter driver 33 determines whether the information processing device 100 is in a normal mode when the information processing device 100 is in a stable state, A high risk mode, which is an access method in a state where there is a high possibility of receiving an impact, is set. Then, the operation mode is switched according to the result of the determination of the state of the information processing apparatus 100 by the shock manager 32.
In the normal mode, the HDD filter driver 33 merely mediates the interface between the HDD driver 36 and the IDE device driver 37, and does not perform any special operation. On the other hand, in the high risk mode, the HDD filter driver 33 (1) subdivides the access unit and (2) invalidates the write cache function for access to the magnetic disk device 10 based on the control of the file system 35. Are added. As a result of this additional operation, when the host computer 30 issues an unload command, the host computer 30 immediately executes the unload command, so that the magnetic head 11 of the magnetic disk device 10 can be evacuated from the magnetic disk 12.

Hereinafter, the operation of the HDD filter driver 33 in the high risk mode will be described in detail.
(1) Subdivision of Access Unit In the magnetic disk device 10, one data file recorded on the magnetic disk 12 is usually an integral multiple of a sector which is a recording unit of the magnetic disk 12, or a minimum recording unit of a file system. A plurality of continuous blocks (data blocks) having a variable length and an integral multiple of a cluster are configured.
As described above, in the general magnetic disk device 10, a response to the host computer 30 cannot be made during a process such as data writing. Therefore, even if the shock manager 32 detects the start of the fall of the information processing apparatus 100 and transmits the unload command from the host computer 30 to the magnetic disk device 10, the unload command is executed until the process is completed and the magnetic head 11 cannot be evacuated. Therefore, when the data file to be accessed includes a large data block (covering many sectors and clusters), it may take a long time before the unload command is executed. Similarly, when the magnetic disk 12 is formatted, the control of the magnetic disk device 10 is occupied for a long time by the operation of the format.

In order to avoid such a situation, in the high risk mode, the HDD filter driver 33 takes a long time to access the data block requested by the file system 35 because the data block to be accessed is a relatively large block. If it is anticipated that the data block (large block) will be required, the data block (large block) is divided into small blocks of a certain size or less, access requests to the data block are subdivided into access requests to individual small blocks, and Access. By setting the size of the small block to a size corresponding to, for example, a unit of a sector or a cluster, the time required for accessing each individual small block can be reduced to the utmost. In the access request requested from the file system 35, the transfer data size, the write destination, and the like are specified, so that the size of the data block is recognized from the contents of the request, and it is determined whether the access takes a long time. be able to. As a result, the data size to be accessed once is reduced, so that the time occupied by the control of the magnetic disk device 10 for the access operation does not exceed a certain length.

FIG. 10 is a diagram illustrating the concept of subdivision of an access unit.
In this case, as shown in FIG. 10, the HDD filter driver 33 notifies the file system 35 that the access to the large block before the subdivision is completed after the access to all the small blocks after the subdivision is completed. I do. As a result, the segmentation operation is hidden from the file system 35.
On the other hand, if an unload command is issued during access to the large block, the unload command is executed when access processing to one small block being executed is completed. Thus, it is possible to prevent the time from issuance to execution of the unload command from being lengthened.
If the magnetic head 11 is unloaded during access to the large block, the small block for which processing has not been completed is accumulated and stored in the internal queue of the HDD filter driver 33. Then, after the magnetic head 11 is loaded, the data transfer request of the small block stored in the internal queue and waiting is discharged in the FIFO format, and is continuously executed.

(2) Invalidation of Write Cache Function As described above, in order to improve the response speed to the host computer 30, the magnetic disk device 10 generally includes a cache memory and performs write cache when writing data. It is. When the magnetic head 11 is retracted from the magnetic disk 12 by unloading, the cache is flushed and the volatile data in the cache memory is replaced with the magnetic data (non-volatile data) on the magnetic disk 12 prior to the execution of the unloading. Data). Therefore, even if the shock manager 32 detects the start of the fall of the information processing device 100 and transmits the unload command from the host computer 30 to the magnetic disk device 10, the cache is flushed and the unload command is executed. It may take a long time.

In order to avoid such a situation, the HDD filter driver 33 disables the write cache function of the magnetic disk device 10 in the high risk mode. Thus, when the unload command is issued, the unload command can be executed immediately without flushing the cache, and the magnetic head 11 can be evacuated from the magnetic disk 12.
When the shock manager 32 determines that the state of the information processing apparatus 100 is stable and the HDD filter driver 33 switches the operation to the normal mode, the write cache function of the magnetic disk device 10 is enabled again (Enable). ).

As a modification of the invalidation of the write cache function described above, the cache is flushed whenever dirty data occurs in the cache memory, so that the cache is substantially in the same state as when the write cache function is invalid (in the write cache). (A state in which dirty data is not stored in the memory).
If the write cache function is completely disabled, the unloading of the magnetic head 11 can be performed quickly, but in a normal access process, the response speed to the host computer 30 by the write cache can be improved. No, performance will be slightly reduced.
Therefore, in this method, the HDD filter driver 33 does not invalidate the write cache function in the high risk mode. Instead, immediately after writing data to the magnetic disk 12, a command to flush the cache (Cache Flush command) is transmitted to the magnetic disk device 10 immediately after the data is written to the magnetic disk 12, and the dirty data generated in the cache memory due to the data writing is transmitted. To delete. Thereby, the cache memory can be used in the data write operation, so that the response speed to the host computer 30 can be improved. On the other hand, dirty data is not stored in the cache memory except when the data write is executed. Can be created with no write cache.
This is because the access to the magnetic disk device 10 does not always take a long time continuously, and therefore, if a cache flush operation is inserted during the idle period of the magnetic disk device 10 between access operations, the access is asynchronous. In many cases when the unloading of the magnetic head 11 occurs, it means that dirty data does not exist in the cache memory. That is, the concept is that an effect equivalent to substantially disabling the write cache function can be obtained.

As described above, when the information processing apparatus 100 is in an unstable state and is likely to receive an impact due to a drop or the like, the HDD filter driver 33 switches to the high risk mode and controls the IDE device driver 37, In accessing the magnetic disk device 10, operations such as (1) subdivision of the access unit and (2) invalidation of the write cache function (including substantial invalidation by immediate flushing of the cache) are performed. As a result, the unloading is performed immediately after the shock is predicted by the shock manager 32, so that even when the fall distance is short (when the time from the start of the fall to the occurrence of the shock is short), the magnetic head 11 is sufficiently placed on the magnetic disk 12. Can be evacuated.
Note that these two additional operations do not depend on each other, and shorten the time until the unloading is executed independently. Therefore, only one of them may be executed.

By the way, once the shock manager 32 determines that the information processing apparatus 100 is highly likely to receive an excessive impact, and once the magnetic head 11 is unloaded, such a high-risk state is released until the state is released. Even if a new access request to the magnetic disk device 10 is transmitted from the file system 35, the HDD filter driver 33 only accepts this access request, and the access to the magnetic disk device 10 is not executed. The case in which the high-risk state is released means that the shock prediction is incorrect as described above in the operation of the evacuation condition calculation unit 77 and the return condition calculation unit 78, and the case of the information processing apparatus 100 that has performed the prediction. If the fluctuation has disappeared and it is determined that the state has returned to a stable state, or a shock has occurred as expected, and the state has been stable for a period sufficient to determine that the shock has passed (time that is considered empirically sufficient). Is determined to be in The file system 35 waits for subsequent processing until the access request is completed (IO Complete). That is, until the information processing apparatus 100 satisfies the condition for returning the magnetic head 11, the magnetic head 11 is not loaded by a new access request from the file system 35.

FIG. 11 is a diagram illustrating the operation of the HDD filter driver 33 in this case.
Whether the magnetic head 11 has already been retracted and the information processing apparatus 100 is in a stationary state or a posture variation state within a certain range equivalent thereto for a certain period (that is, in a stable state), or It is assumed that a data transfer request (access request) from the file system 35 arrives at the HDD filter driver 33 while a mouse / keyboard input by a user interaction for loading a file is not satisfied. . In this case, the data transfer request is accumulated in an internal queue (data transfer request accumulation queue) 331 of the HDD filter driver 33. When the return condition of the magnetic head 11 in the information processing apparatus 100 is satisfied and the return request of the magnetic head 11 is transmitted from the shock manager 32, the HDD filter driver 33 receiving the request first sends the internal queue 331 until then. The data transfer request stored in the buffer is flushed in a FIFO format (queue flush command), and corresponds to the processing of the next newly arrived data transfer request. The data transfer request that arrives during flushing of the internal queue 331 corresponds to such a transient state by being sequentially accumulated in the internal queue 331 even during flushing. Then, after all the data transfer requests accumulated in the internal queue 331 are ejected, the flow returns to the normal flow control of the data transfer request.
By the above operation, the magnetic head 11 is inadvertently returned to the original state during the unloading of the magnetic head 11, and the magnetic head 11 collides with the magnetic disk 12 due to a subsequent external impact, thereby damaging the magnetic disk 12. Thus, it is possible to avoid a situation where some or all of the data cannot be restored.

In the case where the access unit is subdivided, if the unloading of the magnetic head 11 is performed at a timing when all of the subdivided access requests have not yet completed the processing, the saved magnetic head 11 is restored until the magnetic head 11 is restored. The transmission of the remaining subdivided access requests to the IDE device driver 37 remains suspended (Suspended). Therefore, even in this case, it is possible to prevent the magnetic head 11 from returning to the original position and damaging the magnetic disk 12.

(4) When the return condition of the magnetic head 11 is satisfied, an operation request (load command) necessary for loading the magnetic head 11 and returning it to the magnetic disk 12 is sent from the HDD filter driver 33 to the IDE device driver 37. However, when the magnetic disk device 10 automatically loads the magnetic head 11 by sending a normal data processing request, such an operation request is unnecessary. However, when it is desired to reduce or omit the time required for returning the magnetic head 11 at the time of accessing data in the next magnetic disk device 10, the return of the magnetic head 11 is automatically performed along with the data access. However, when the return condition of the magnetic head 11 is satisfied, a return operation request (load command) may be sent regardless of the presence or absence of the data access request.

In the above-described embodiment, the HDD driver 36, the HDD filter driver 33, and the IDE device driver 37 have been described as separate modules. However, in terms of mounting, they are configured as a single module including all the functions of these drivers. It is also possible.

Further, in the present embodiment, the unload command is used to retreat the magnetic head 11 from above the magnetic disk 12, but a magnetic disk device of a specification in which there is no dedicated command for unloading the magnetic head 11 exists. For 10, a power saving command (standby command or the like) for stopping the spindle motor or the like for driving the magnetic disk 12 to rotate can also be used.

By the way, according to the above-described embodiment, even when the information processing apparatus 100 receives an impact, the magnetic head 11 cannot be retracted in time and a failure occurs in the magnetic disk device 10 to cause an operation failure. It cannot be said that it can be prevented. In the present embodiment, the inclination and vibration of the information processing apparatus 100 are analyzed, and the magnetic head 11 is retracted after the state of the information processing apparatus 100 shifts from the first stage to the second stage in FIG. Therefore, in the case where the time from the transition to the second stage to the occurrence of an impact is extremely short, such as when the information processing apparatus 100 in use (in a stationary state) being used on a desk is lifted to be lifted. However, an impact may occur before the retraction operation of the magnetic head 11 according to the present embodiment is completed, and the magnetic disk 12 may be damaged.

(4) When a permanent defective portion (scratch) occurs in the magnetic recording area of the magnetic disk 12 due to the impact received by the magnetic disk device 10, the recording data in the area is directly lost. Furthermore, when the hard magnetic head 11 collides with the surface of the magnetic disk 12 and fine fragments of the coated magnetic surface, which have been scraped off, scatter inside the magnetic disk device 10, the device normally operates normally. There is a risk that such fragments may be caught in the gap between the magnetic head 11 and the magnetic disk 12 and cause a secondary obstacle such as making a larger scratch on the surface of the magnetic disk 12.

If it is found with high accuracy that such a defect, that is, a defective portion due to physical scratches has occurred in the magnetic recording area of the magnetic disk 12, the user can interrupt the work at that time and be affected by the defect. You will want to know the degree in detail. In addition, in order to minimize damage in the near future, it is preferable to immediately perform a recovery operation such as taking a data backup.
Therefore, as another embodiment of the present invention, it is checked whether the retraction of the magnetic head 11 is completed before the shock is actually generated after the shock generated in the magnetic disk device 10 is predicted as described above. In this way, it is possible to confirm whether the magnetic disk 12 has been damaged by the impact and to propose a means for notifying the user of the necessity of taking measures such as data backup.

In the protection mechanism of the magnetic disk device 10 according to the present embodiment, the time from the start of the retreat operation of the magnetic head 11 to the completion thereof and the start of the retreat operation are added to the shock manager 32 of the protection mechanism shown in FIG. And a function of measuring and comparing the time from when the magnetic head 11 is retracted to the time when the magnetic head 11 is retracted is added.
FIG. 12 is a diagram illustrating a functional configuration of the shock manager 40 according to the present embodiment.
As shown in FIG. 12, the shock manager 40 includes a sensor monitor 61, an acceleration data history storage 62, a timer drive controller 63, a keyboard / mouse event history storage 64, which constitute the shock manager 32 shown in FIG. In addition to the shock prediction unit 65, as means for diagnosing whether or not the magnetic head 11 has been retracted before the occurrence of a shock, an impact grace period measuring unit 121, a magnetic head withdrawal completion time measuring unit 122, a diagnostic processing unit 123, and a magnetic head A position confirmation unit 124 is provided.

When the retracting operation of the magnetic head 11 is performed, the shock grace period measuring unit 121 measures the time (shock grace period) from the start time to the time when the shock actually occurs. As shown in FIG. 12, the shock delay period measuring unit 121 receives a notification of the start of the retreat operation of the magnetic head 11 from the HDD filter driver 33 and records the input time. In addition, acceleration data is input from the sensor monitor 61, and the time of occurrence of the impact is recorded.
As shown in the drawing, the shock grace period measuring unit 121 includes a high-strength shock estimating unit 125. The high-strength shock estimating unit 125 analyzes the acceleration data input from the sensor monitor 61 to determine the occurrence of a shock. Detect. In other words, the high-strength impact estimation unit 125 detects a characteristic impact detected by the sensor 20 when a severe impact (high-strength impact) exceeding the measurable range of the sensor 20 occurs due to a drop of the information processing apparatus 100 or the like. Understand the impact amplitude pattern and identify its high intensity impact. Then, the time when the occurrence of the high strength impact is detected is obtained.
The shock grace period measuring unit 121 calculates the shock grace period from the start time of the retracting operation of the magnetic head 11 specified by the notification from the HDD filter driver 33 and the detection time of the high strength shock by the high strength shock estimation unit 125. Is calculated and sent to the diagnosis processing unit 123.

The magnetic head retraction completion time measuring unit 122 measures the time required for completing the retraction operation when the retraction operation of the magnetic head 11 is performed. As shown in FIG. 12, the magnetic head evacuation completion time measuring unit 122 receives a start notification and a completion notification of the evacuation operation of the magnetic head 11 from the HDD filter driver 33, and calculates the time required for the evacuation operation from the input time. I do. In order to reduce the load on the shock manager 40 and the HDD filter driver 33, instead of measuring the time actually required when the magnetic head 11 is evacuated each time, the time actually measured at the time of booting or detection of the magnetic disk device 10 is actually measured in advance. The evacuation operation may be performed, the evacuation time may be measured and stored in the cache memory, and the evacuation operation may be used as the time required for completing the evacuation operation at the time of impact prediction. Furthermore, a unique value evaluated in advance based on the specifications of the magnetic disk device 10 or the like may be stored in a memory such as a ROM and used as a time required for completing the evacuation operation at the time of shock prediction.
The magnetic head withdrawal completion time measuring unit 122 sends the withdrawal time of the magnetic head 11 at the time of impact prediction obtained as described above to the diagnostic processing unit 123.

The diagnosis processing unit 123 retreats the magnetic head 11 based on the shock grace period received from the shock grace period measuring unit 121 and the evacuation time of the magnetic head 11 at the time of shock prediction received from the magnetic head evacuation completion time measuring unit 122. It is determined whether or not the operation was in time by the time of the actual impact. Then, when it is determined that the magnetic head 11 has not been in time, a warning message indicating that the magnetic head 11 may have collided with the magnetic disk 12 and caused a damage is output as a diagnosis result and notified to the user.

FIG. 13 is a diagram for explaining the relationship between the shock delay period used for the determination by the diagnosis processing unit 123 and the evacuation time of the magnetic head 11.
In the present embodiment, as shown in FIG. 4, the process from when the information processing apparatus 100 starts swinging to when the information processing apparatus 100 falls and generates an impact is classified into four stages. When it is determined that the state has shifted from the first stage to the second stage, the evacuation request of the magnetic head 11 is issued. Therefore, as shown in FIGS. 13A and 13B, an impact (an impact having a strength exceeding the limit (maximum allowable shock range) that the magnetic disk device 10 can withstand) actually occurs from the time when the evacuation request of the magnetic head 11 is issued. There is some time lapse before. This is the shock grace period measured by the shock grace period measuring unit 121, and the diagnosis processing unit 123 verifies whether the retraction of the magnetic head 11 has been completed within this shock grace period.

As shown in FIG. 13A, when the retraction time of the magnetic head 11 is shorter than the shock grace period, it is determined that the retraction of the magnetic head 11 is completed before the impact occurs, and the magnetic recording area of the magnetic disk 12 is determined. A diagnostic result that no defective area has occurred is obtained. On the other hand, as shown in FIG. 13 (B), when the retraction time of the magnetic head 11 is longer than the impact delay period, it is determined that an impact has occurred before the retraction of the magnetic head 11 is completed, and the magnetic disk 12 A diagnosis result is obtained that a defective area may have occurred in the magnetic recording area.
The diagnosis result may be notified to the user only when the retraction of the magnetic head 11 has not been completed by the time of the impact as described above, or when the retraction of the magnetic head 11 is completed by the time of the impact. The user may be notified that there is no possibility of occurrence of a defective area. Further, the warning message output as the diagnosis result may be a content that prompts the user to immediately take immediate action such as data backup.

Here, the shock grace period measuring unit 121 calculates the shock grace period from the start time of the retracting operation of the magnetic head 11 and the shock occurrence time, and the magnetic head evacuating completion time measuring unit 122 calculates the magnetic head 11 at the time of shock prediction. The evacuation time is obtained and compared, but other implementations are possible. For example, the time when the occurrence of the high-intensity impact is detected by the high-intensity impact estimating unit 125 is directly input to the diagnosis processing unit 123, and the magnetic head retraction completion time measuring unit 122 starts the retreat operation of the magnetic head 11 at the start time. And the evacuation completion time calculated from the pre-measured evacuation time is input to the diagnostic processing unit 123. Based on the time after the evacuation completion time, whether the evacuation operation of the magnetic head 11 has been completed by the time of the impact is generated. It is also possible to diagnose whether or not it is not.

When the shock prediction unit 65 sends the HDD filter driver 33 a retreat request for the magnetic head 11, the magnetic head position checking unit 124 checks whether the magnetic head 11 is on the magnetic disk 12 Check if it is far away and in the evacuation position. That is, the magnetic head 11 may be unloaded irrespective of the impact prediction due to the power saving function or the like. If the magnetic head 11 has been unloaded at the time when the impact prediction unit 65 sends the evacuation request for the magnetic head 11, there is no need to perform a subsequent diagnosis by the diagnosis processing unit 123. The position of the head 11 (load or unload) is verified.

Here, when the HDD filter driver 33 issues an unload command in response to the evacuation request transmitted from the shock prediction unit 65 and the magnetic head 11 is evicted, a notification of evacuation completion is returned from the controller of the magnetic disk device 10. come. Therefore, if the magnetic head 11 has already been unloaded at the time of starting the operation, it takes an extremely short time (usually several hundred msec, but several tens to 100 μsec (microsecond)) as compared with the normal operation. The operation completion notification is returned. Therefore, when an operation completion notification is returned within a predetermined time (about 100 μsec) after the HDD filter driver 33 issues the unload command, the operation completion notification is sent from the HDD filter driver 33 to the magnetic head position confirmation unit 124. I will send it. Then, based on this notification, the magnetic head position confirmation unit 124 controls the diagnosis processing unit 123 to end the diagnosis processing.

FIG. 14 is a flowchart illustrating a method of confirming the position of the magnetic head 11 by the HDD filter driver 33 and the magnetic head position confirmation unit 124.
First, the HDD filter driver 33 issues an unload command in response to the evacuation request from the shock prediction unit 65 (step 1401), and measures the elapsed time from this point (step 1402). If the evacuation processing of the magnetic head 11 is completed before the elapsed time exceeds a preset time T (for example, 100 μsec) (step 1403), the HDD filter driver 33 notifies the operation completion notification of the magnetic head position. This is sent to the confirmation unit 124 (step 1404). Upon receiving this notification, the magnetic head position confirmation unit 124 determines that the magnetic head 11 is already in the unloaded state (step 1405), and terminates the diagnostic processing by the diagnostic processing unit 123 (step 1406).
On the other hand, if the elapsed time after the issuance of the unload command exceeds the time T before the evacuation processing of the magnetic head 11 is completed (steps 1402 and 1403), the HDD filter driver 33 operates the magnetic head position confirmation unit 124. No completion notification is sent. Therefore, the magnetic head position confirmation unit 124 does not determine that the magnetic head 11 is in the unloaded state, and the diagnosis processing by the diagnosis processing unit 123 is continued.

Next, the overall flow of the diagnostic processing after impact prediction according to the present embodiment will be described.
FIG. 15 is a flowchart illustrating the flow of the diagnostic processing according to the present embodiment.
As shown in FIG. 15, when the impact is predicted by the impact predicting unit 65 (steps 1501 and 1502), the HDD filter driver 33 unlocks the magnetic disk device 10 according to the retraction request of the impact predicting unit 65. The load command is issued, and the retreat operation of the magnetic head 11 is started (step 1503).
Then, when the control of the diagnosis processing unit 123 is performed by the magnetic head position confirmation unit 124, that is, when the magnetic head 11 has already been unloaded, the diagnosis processing ends as it is (steps 1504 and 1501). Thus, if the magnetic head 11 has already been unloaded when the retreat operation of the magnetic head 11 is started, the present embodiment can be performed without determining whether the magnetic head 11 has been retreated within the shock delay period. The diagnosis process according to the mode is ended.

When the magnetic head 11 is being loaded and the control of the diagnosis processing unit 123 by the magnetic head position confirmation unit 124 is not performed, the shock delay period measurement unit 121, the magnetic head retraction completion time measurement unit 122, and the diagnosis processing unit 123 Then, it is determined whether or not an impact is detected before the retraction of the magnetic head 11 is completed (steps 1505 and 1506).
If the retracting operation of the magnetic head 11 is completed before the impact is detected, there is no possibility that the magnetic head 11 collides with the magnetic disk 12 and damages the magnetic disk 12, and the diagnostic processing is terminated as it is (steps 1506 and 1501). ).
On the other hand, if an impact is detected before the retracting operation of the magnetic head 11 is completed, it is possible that the magnetic head 11 has collided with the magnetic disk 12 and caused damage, and the user is notified of a warning message. (Steps 1505 and 1507), the diagnostic processing ends (Step 1501).

As described above, according to the diagnostic processing of the magnetic disk device 10 according to the present embodiment, when an impact actually occurs after the impact is predicted, it is determined whether or not the retracting operation of the magnetic head 11 is in time and the magnetic head 11 If there is a possibility that the user has collided with the magnetic disk 12 and caused damage, a warning to that effect is given to the user, and the execution of recovery processing such as data backup can be urged. That is, when it is doubtful whether the magnetic disk device 10 or the information processing device 100 receives a shock and a failure occurs in the operating magnetic disk device 10, the determination can be made with certain accuracy.
Thereby, the fragments of the applied magnetic surface that are scraped off when the magnetic head 11 collides with the magnetic disk 12 may further damage the surface of the magnetic disk 12 during the subsequent operation of the magnetic disk device 10. The failure makes it possible to avoid damage such as loss of more data.

Although the above embodiments have been described assuming the information processing apparatus 100 such as a notebook computer as a target, various embodiments in which a magnetic disk device is mounted as a storage unit, such as a single magnetic disk device or a hard disk recorder, are described. It is needless to say that the present invention can be applied to the above device.

FIG. 1 is a diagram illustrating a configuration of an information processing apparatus to which a protection mechanism of a magnetic disk device according to an embodiment is applied. FIG. 1 is a diagram illustrating an appearance of an information processing apparatus to which a protection mechanism of a magnetic disk device according to an embodiment is applied. FIG. 1 is a diagram schematically illustrating a general device configuration of a magnetic disk device. It is the figure which showed typically the state change from the stable state to the fall of the information processing apparatus. FIG. 2 is a diagram illustrating a system configuration of a protection mechanism of the magnetic disk device according to the embodiment. FIG. 3 is a diagram illustrating a functional configuration of a shock manager according to the present embodiment. FIG. 7 is a diagram illustrating the internal function of the shock prediction unit shown in FIG. 6 in further detail. FIG. 8 is a diagram illustrating an example of an operation algorithm performed by the evacuation condition calculation unit illustrated in FIG. 7. FIG. 8 is a diagram illustrating an example of an operation algorithm by a return condition calculation unit illustrated in FIG. 7. FIG. 3 is a diagram illustrating the concept of subdivision of an access unit according to the present embodiment. FIG. 7 is a diagram illustrating an operation when the HDD filter driver according to the present embodiment receives an access request while the magnetic head is being retracted. It is a figure showing functional composition of a shock manager in other embodiments of the present invention. FIG. 4 is a diagram illustrating a relationship between an impact delay period used for determination by a diagnostic processing unit and a magnetic head retreat time according to the present embodiment. 5 is a flowchart illustrating a magnetic head position confirmation method by the HDD filter driver and the magnetic head position confirmation unit according to the present embodiment. 5 is a flowchart illustrating a flow of a diagnosis process according to the present embodiment.

Explanation of reference numerals

DESCRIPTION OF SYMBOLS 10 ... Magnetic disk drive, 20 ... Sensor, 30 ... Host computer, 31 ... Sensor driver, 32, 40 ... Shock manager, 33 ... HDD filter driver, 34 ... Application, 35 ... File system, 36 ... HDD driver, 37 ... IDE Device driver, 61: Sensor monitor unit, 62: Acceleration data history storage unit, 63: Timer drive control unit, 64: Keyboard / mouse event history storage unit, 65: Impact prediction unit, 71: Previous variation average calculation unit, 72 ... Angle calculator 73, angular velocity calculator 74, fluctuation period detector 75, free fall detector 76, shock detector 77, evacuation condition calculator 78, return condition calculator 121, shock delay period measurement Unit, 122: magnetic head retraction completion time measuring unit, 123: diagnostic processing unit, 124: magnetic head position Checking unit, 125 ... high strength impact estimator

Claims (33)

Information acquisition means for acquiring information on environmental fluctuations with respect to the magnetic disk device;
Shock prediction means for analyzing the information acquired by the information acquisition means together with the history thereof, and judging a state where the magnetic disk device is placed to perform a shock prediction;
Control means for controlling the operation of the magnetic disk drive, including retraction of the magnetic head, based on a result of the prediction by the shock predicting means. 2. The apparatus according to claim 1, wherein the shock prediction means does not predict that a shock will occur depending on the change in the state of the magnetic disk device if the change in the state of the magnetic disk device falls within a certain range for a certain period. 2. A protection mechanism for a magnetic disk according to claim 1. 2. The magnetic disk drive according to claim 1, wherein the shock prediction unit predicts that a shock will occur due to the change in the state when the state of the magnetic disk device is changed in a predetermined pattern. Protection mechanism. 2. The magnetic disk protection mechanism according to claim 1, wherein the impact prediction unit performs an impact prediction by referring to a history of input operations performed by a predetermined input device. The information obtaining means obtains information on acceleration generated in the magnetic disk device,
2. The magnetic disk protection mechanism according to claim 1, wherein the shock predicting unit recognizes a state of the magnetic disk device in which the magnetic disk device is placed, based on the information on the acceleration acquired by the information acquiring unit. . The impact prediction means, when determining that the magnetic disk device is in a stable state, notifies the control means to that effect,
2. The magnetic disk protection mechanism according to claim 1, wherein the control unit returns the retracted magnetic head in response to the notification. The shock predicting means adaptively determines whether or not the magnetic disk device is in a stable state based on a history of information obtained from the information obtaining means before predicting that a shock will occur. 7. The protection mechanism for a magnetic disk according to claim 6, wherein: When the magnetic head is retracted, the control unit does not execute a new access request to the magnetic disk device until the impact prediction unit determines that the magnetic disk device is in a stable state. The magnetic disk protection mechanism according to claim 1, wherein the magnetic disk is held in an internal queue. State determining means for determining a state of the magnetic disk drive;
Control means for controlling the operation of the magnetic disk device including retraction of a magnetic head based on a result of the judgment by the state judgment means,
The control unit, when the state determination unit determines that the magnetic disk device is in a state where there is a high possibility that the magnetic disk device is subjected to an excessive shock, sets an access request to the magnetic disk device as a target of one access. A protection request for a magnetic disk, wherein the request is divided into access requests having a small data size and transmitted to the magnetic disk device. When the magnetic head is evacuated before at least a part of the subdivided access request is executed, the control unit saves the access request that has not been executed and returns the magnetic head. The protection mechanism of a magnetic disk according to claim 9, wherein the protection mechanism is executed later. State determining means for determining a state of the magnetic disk drive;
Control means for controlling the operation of the magnetic disk device including retraction of a magnetic head based on a result of the judgment by the state judgment means,
The control means invalidates the write cache function of the magnetic disk device when the state determination means determines that the magnetic disk device is in a state in which the magnetic disk device is highly likely to be subjected to an excessive impact. A protection mechanism for a magnetic disk, characterized by accessing the disk. State determining means for determining a state of the magnetic disk drive;
Control means for controlling the operation of the magnetic disk device including retraction of a magnetic head based on a result of the judgment by the state judgment means,
The control unit is configured to, when the state determination unit determines that the magnetic disk device is in a state in which the magnetic disk device is likely to receive an excessive impact, every time write data is generated in the cache memory included in the magnetic disk device. An operation of writing the write data to the magnetic disk to empty the cache memory. Shock prediction means for predicting a shock that can occur on the magnetic disk device based on a change in environment with respect to the magnetic disk device;
Control means for controlling the operation of the magnetic disk device including retraction of a magnetic head based on a result of prediction by the impact prediction means;
Diagnosing means for determining whether or not the magnetic head has been completely retracted before the occurrence of an impact when the impact actually occurs after the magnetic head is retracted by the control means. Magnetic disk protection mechanism. The diagnostic means compares the shock grace period, which is the time from the start of the retracting operation of the magnetic head to the time of occurrence of the shock, with the retracting time required for the retracting operation of the magnetic head, which is measured and held in advance, 14. The protection mechanism for a magnetic disk according to claim 13, wherein the determination is performed. If the magnetic head has already been retracted when the magnetic head retreat operation is started under the control of the control means, the diagnostic means does not compare the shock grace period with the retreat time, and 15. The magnetic disk protection mechanism according to claim 14, wherein it is determined that the retraction of the magnetic head has been completed before the occurrence of the error. The diagnostic means, when the control means issues a request command for controlling the magnetic disk device to perform the evacuation operation of the magnetic head, and obtains a completion notification of the command within a predetermined time, 16. The magnetic disk protection mechanism according to claim 15, wherein it is determined that the magnetic head has already been retracted when the magnetic head retract operation is started. In a computer system having a magnetic disk device,
An acceleration sensor for detecting acceleration generated in the housing;
A shock manager that analyzes information of acceleration and its history acquired by the acceleration sensor and predicts occurrence of an impact on the housing,
A computer system comprising: a driver that controls an operation of the magnetic disk device including a retraction of a magnetic head based on a result of prediction by the shock manager. In a computer system having a magnetic disk device,
A shock manager for judging the state of the housing and predicting the occurrence of an impact on the housing;
If the shock manager determines that the magnetic disk device is in a state where there is a high possibility that the magnetic disk device will be subjected to an excessive shock, an access request to the magnetic disk device is issued with a small data size for one access. And a driver for subdividing the access request and transmitting the access request to the magnetic disk device. In a computer system having a magnetic disk device,
A shock manager for judging the state of the housing and predicting the occurrence of an impact on the housing;
A driver for accessing the magnetic disk by disabling a write cache function provided in the magnetic disk device when the shock manager determines that the magnetic disk device is in a state of high possibility of receiving an excessive shock; A computer system comprising: In a computer system having a magnetic disk device,
An acceleration sensor for detecting acceleration generated in the housing;
An impact prediction unit that analyzes the information and the history of the acceleration obtained by the acceleration sensor and predicts the occurrence of an impact on the housing,
A driver that controls an operation of the magnetic disk device including a retraction of a magnetic head based on a prediction result by the shock prediction unit;
A diagnostic processing unit that, when an impact actually occurs after the driver starts retreating the magnetic head, determines whether or not the retraction of the magnetic head is completed before the impact occurs. Computer system. If the diagnostic processing unit determines that an impact has occurred before the retraction of the magnetic head is completed, the diagnostic processing unit notifies the user of a warning that a failure may have occurred in the magnetic disk device. 21. The computer system according to claim 20, wherein: In a magnetic disk protection method for determining a placed state of a magnetic disk device using a sensor and retracting a magnetic head according to the determination result to protect the magnetic disk,
Accumulate the history of information obtained from the sensor,
Analyzing the accumulated history and the last obtained information to recognize the variation pattern of the state of the magnetic disk device,
A method for protecting a magnetic disk, comprising: retracting the magnetic head when it is predicted that an impact will occur on the magnetic disk device based on the content of a change in the state of the magnetic disk device. In a magnetic disk protection method for determining a placed state of a magnetic disk device using a sensor and retracting a magnetic head according to the determination result to protect the magnetic disk,
Determining a put state of the magnetic disk device based on an output of the sensor;
When it is determined that the magnetic disk device is in a state where there is a high possibility of receiving an excessive impact, an access request to the magnetic disk device is subdivided into an access request having a small data size to be accessed once. Performing operation control so that the data is transmitted to the magnetic disk device.
Retreating the magnetic head when the magnetic disk device is predicted to receive an excessive impact. In a magnetic disk protection method for determining a placed state of a magnetic disk device using a sensor and retracting a magnetic head according to the determination result to protect the magnetic disk,
Determining a put state of the magnetic disk device based on an output of the sensor;
When it is determined that the magnetic disk device is in a state where there is a high possibility of receiving an excessive impact, the operation control is performed such that the write cache function of the magnetic disk device is invalidated and the magnetic disk is accessed. Steps to perform;
Retreating the magnetic head when the magnetic disk device is predicted to receive an excessive impact. In a magnetic disk protection method for determining a placed state of a magnetic disk device using a sensor and retracting a magnetic head according to the determination result to protect the magnetic disk,
Determining a put state of the magnetic disk device based on an output of the sensor;
When it is determined that the magnetic disk device is in a state where it is highly likely to receive an excessive shock, each time write data is generated in the cache memory provided in the magnetic disk device, the write data is written to the magnetic disk. Performing an operation control to perform an operation of writing and emptying the cache memory;
Retreating the magnetic head when the magnetic disk device is predicted to receive an excessive impact. In a magnetic disk protection method for determining a placed state of a magnetic disk device using a sensor and retracting a magnetic head according to the determination result to protect the magnetic disk,
Estimating a shock that may occur on the magnetic disk device based on a change in environment with respect to the magnetic disk device;
Controlling the operation of the magnetic disk drive including retracting the magnetic head based on the result of the prediction;
If an impact actually occurs after the magnetic head retreat starts, an impact grace period, which is the time from the start of the retraction operation of the magnetic head to the time of the impact, and the retraction of the previously measured and held magnetic head Determining whether or not the retraction of the magnetic head has been completed before the occurrence of the impact by comparing the retraction time required for the operation. In the step of determining whether the retraction of the magnetic head has been completed before the occurrence of the impact, if the magnetic head has already been retracted when the retreat operation of the magnetic head is started, the impact grace period and the 27. The magnetic disk protection method according to claim 26, wherein it is determined that the retraction of the magnetic head has been completed before the occurrence of the impact without comparing with the retraction time. A program for controlling a computer to realize a protection mechanism of a magnetic disk device,
Means for acquiring information on environmental fluctuations with respect to the magnetic disk device and storing the information in predetermined storage means;
Means for analyzing the acquired information and the history of the information stored in the storage means, and determining the state of the magnetic disk device to perform impact prediction;
A computer-readable storage medium storing a program for causing a computer to function as means for controlling an operation of the magnetic disk device including a retraction of a magnetic head based on a result of the shock prediction. A program for controlling a computer to realize a protection mechanism of a magnetic disk device,
Means for determining the state of the magnetic disk drive;
Means for acquiring information on environmental fluctuations with respect to the magnetic disk device and storing the information in predetermined storage means;
When it is determined that the magnetic disk device is in a state where there is a high possibility of receiving an excessive impact, an access request to the magnetic disk device is subdivided into an access request having a small data size to be accessed once. Means for controlling the operation so as to perform an operation of converting the data to the magnetic disk device,
A computer-readable storage medium storing a program for causing a computer to function as a unit for retracting a magnetic head when the magnetic disk device is predicted to receive an excessive impact. A program for controlling a computer to realize a protection mechanism of a magnetic disk device,
Means for determining the state of the magnetic disk drive;
Means for acquiring information on environmental fluctuations with respect to the magnetic disk device and storing the information in predetermined storage means;
When it is determined that the magnetic disk device is in a state where there is a high possibility of receiving an excessive impact, the operation control is performed such that the write cache function of the magnetic disk device is invalidated and the magnetic disk is accessed. Means to do,
A computer-readable storage medium storing a program for causing a computer to function as a unit for retracting a magnetic head when the magnetic disk device is predicted to receive an excessive impact. A program for controlling a computer to realize a protection mechanism of a magnetic disk device,
Means for determining the state of the magnetic disk drive;
Means for acquiring information on environmental fluctuations with respect to the magnetic disk device and storing the information in predetermined storage means;
When it is determined that the magnetic disk device is likely to be subjected to excessive shock, each time write data is generated in the cache memory provided in the magnetic disk device, the write data is written to the magnetic disk. Means for performing operation control so as to perform an operation of emptying the cache memory,
A computer-readable storage medium storing a program for causing a computer to function as a unit for retracting a magnetic head when the magnetic disk device is predicted to receive an excessive impact. A program for controlling a computer to realize a protection mechanism of a magnetic disk device,
Means for predicting an impact that may occur on the magnetic disk device based on a change in environment with respect to the magnetic disk device;
Means for controlling the operation of the magnetic disk device, including retracting the magnetic head, based on the result of the prediction;
If an impact actually occurs after the magnetic head retreat starts, an impact grace period, which is the time from the start of the retraction operation of the magnetic head to the time of the impact, and the retraction of the previously measured and held magnetic head By comparing the evacuation time required for the operation with the evacuation time, as means for determining whether or not the evacuation of the magnetic head has been completed before the occurrence of the impact,
A program for causing the computer to function. A computer-readable recording medium storing the program according to any one of claims 28 to 32.

Images