JP2001325048A - Disk storage device and power-managing method in the same device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize optimal power saving control by switching and setting an operation threshold for power control, immediately after the generation of events accompanied with the start of this system and after the lapse of a fixed time which elapses since then. SOLUTION: A threshold for power control is set as t1 immediately after the generation of specific events accompanying the start of this system, such as power supply (a step S1), and in a fixed time tp from the point of time (a step S3), power saving control is performed by using the t1 (a step S2). After the lapse of the fixed time tp, the threshold for power control is switched and set from the t1 to t2 (a step S4), and the power saving control is performed, by using the t2 (a step S5) until the power source is disconnected (a step S6).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention sets a predetermined portion in a device to a power saving state when an access request from a host is interrupted for a time longer than a time specified by a power management threshold value. The present invention relates to a disk storage device that performs power management to return to the original state when an access request from the host occurs in the state, and in particular, for a certain period of time immediately after the occurrence of a specific event accompanying system startup, such as power-on, and the like. The present invention relates to a disk storage device suitable for performing optimal power management after a lapse of a predetermined time and a power management method in the device.

[0002]

2. Description of the Related Art In portable personal computers and the like,
From the viewpoint of power saving, finely managing power according to the state has been widely adopted. Even in a magnetic disk device, which is a main element of a personal computer, power of a read / write circuit system, a positioning circuit system, a spindle motor drive circuit system, and the like can be managed in accordance with an access state from a host (host system). In the magnetic disk device having the power management function (power management mechanism), when the access from the host is interrupted, (1) disconnection of the read / write circuit system (electrical disconnection) at every predetermined elapsed time (2) Opening of the positioning circuit system (head retract state) (3) The power management mechanism operates in the order of stopping the rotation of the spindle motor, and gradually shifts from a weak power saving state to a strong power saving state.

When an access request is issued from the host, the power management mechanism returns from the above power saving state to a steady state, and performs an operation according to the access request.

[0004] Therefore, if an access request is made while the power management mechanism is in the power saving state, the time until recovery is a delay in the operation for the access request.

[0005]

As described above, the power management mechanism of the magnetic disk drive (power management mechanism) performs power management when access from the host is interrupted based on a predetermined elapsed time (operation threshold). Have gone. In a conventional magnetic disk drive, the operation threshold time is determined uniformly.

On the other hand, a personal computer (hereinafter, referred to as a PC) using a magnetic disk device built in or connected externally.
), The frequency of access to the magnetic disk device differs depending on the use state. However, during a system startup period such as when the PC is turned on, the access frequency distribution is determined regardless of the use state.

Therefore, if the power management (operation threshold value used) by the power management mechanism is optimized based on the system startup state, the power management in the normal state after the system startup is not optimized. If optimization is performed on the basis of the normal state, a contradiction occurs that optimization is not performed when the system is started.

For example, in the case of a portable PC, there is a strict requirement for the life of the battery or the charging interval. In a portable PC, reading and creating texts are mainly used, so block data transfer continues at short access intervals when reading or writing to a disk. After that, since the user performs an operation of reading and writing the sentence displayed on the screen as a target, the access request pause continues for a long time.

The optimum power management condition of the power management mechanism for such a use state is to set the power saving state as soon as the access request is interrupted. However, if optimization is performed for reading and writing ordinary texts, the power saving function of the power management mechanism operates when the system is started up, and there is a problem that it takes time to start up.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and has as its object the operation threshold for power management immediately after the occurrence of a specific event, such as power-on, which accompanies system startup, and after a certain period of time has elapsed. It is an object of the present invention to provide a disk storage device and a power management method in the same device, which can always realize optimal power-saving management while minimizing an operation delay for an access request from a host by changing the setting.

Another object of the present invention is to dynamically change and set the threshold according to the use state of the apparatus after a lapse of a predetermined time from the occurrence of the specific event, so that the normal state after system startup is established. It is an object of the present invention to provide a disk storage device and a power management method in the disk device, which can realize more optimal power saving management while suppressing a decrease in performance in the device.

Still another object of the present invention is to use a threshold determined independently of the dynamically changed threshold for a certain period of time after the occurrence of the specific event, so that the system can be used at system startup. It is an object of the present invention to provide a disk storage device capable of more effective power saving management without sacrificing system startup time and a power management method in the device.

[0013]

According to the present invention, when an access request from a host is interrupted for a time longer than a time specified by a power management threshold, a predetermined portion in the device is set to a power saving state. Power management means for setting and returning to the original state when an access request from the host occurs in the power saving state, and setting the power management threshold value to a value immediately after the occurrence of a specific event involving system startup, and then to a certain value. Threshold switching setting means for switching and setting after a lapse of time.

In the disk storage device having such a configuration, the power management threshold is switched between immediately after the occurrence of a specific event accompanying system startup and after a certain period of time has elapsed.

As already described in the section of [Problems to be Solved by the Invention], the use state such as the access frequency of the disk storage device differs between when the system is started and after the system is started. Therefore, instead of determining the power management threshold (operation threshold) based on either the system startup state or the state after the system startup as in the related art, a system according to the present invention is used. By switching and using the above threshold immediately after the occurrence of a specific event accompanied by start-up and after a certain period of time has elapsed, the system can be used in both a normal use state (normal state) at system start-up and after system start-up. In addition, optimum power management (power saving management) can be realized without causing unnecessary operation delay in response to an access request from the host.

The access frequency from the host to the disk storage device is determined at the time of system start-up irrespective of the use state, as described in the section "Problems to be Solved by the Invention". In the normal state after raising, it differs depending on the use state.

Therefore, in the normal state after system startup, that is, after a certain period of time has elapsed after the occurrence of a specific event accompanying system startup, the threshold value is set to the usage state of the disk storage device, for example, from the host to the disk storage device. It is good to dynamically switch and set according to the frequency of access to.

In this way, power saving management in a normal state after system startup can be performed more optimally while suppressing a decrease in performance.

Further, by setting the power management means to prevent a predetermined portion in the apparatus from being set to the power saving state for a certain period immediately after the occurrence of the specific event, the power management means ( The power saving function of the power management mechanism works, and it is possible to prevent the occurrence of a problem that it takes time to start up. Here, it is possible to independently prepare means for suppressing the setting of the power saving state, but it is also possible to realize the suppressing means by the threshold value switching setting means. For this purpose, the threshold value for power management set by the threshold value switching setting means immediately after the occurrence of the specific event may be set to a value larger than the above-mentioned fixed time. Obviously, in this case, the power management unit does not operate for a certain period of time immediately after the occurrence of the specific event, and therefore, the power management unit is prevented from being set to the power saving state.

After a certain period of time has elapsed immediately after the occurrence of a specific event, even if the above-mentioned threshold value is dynamically switched and set according to the use state of the disk storage device, for example, once the power is turned off, Then, when the above-mentioned specific event occurs again, such as when the power is turned on again, the threshold value that was set most recently, that is, the threshold value that was set last in the normal state after system startup Is not always optimal power-saving management because the usage state at the time of system startup differs from the usage state after system startup. Therefore, for a certain period immediately after the occurrence of the specific event,
If a threshold value determined independently of the dynamically set threshold value is used, it is possible to prevent a problem that the power management means operates unnecessarily at the time of starting up the system and takes a long time to start up the system.

Since the use state of the disk storage device is determined by the frequency of access requests from the outside (for example, a host), a threshold value specified from the outside is set after a lapse of a predetermined time immediately after the occurrence of the specific event. The power saving management is performed, and a certain time immediately after the occurrence of the specific event is set to a value larger than the certain time as the threshold, or a threshold determined independently of the threshold specified from the outside. It is good to set. In this way, more effective power saving management can be achieved without sacrificing the system startup time at the time of system startup and while minimizing the operation delay for access requests from the host in the normal state after system startup. Can be realized.

The above-mentioned specific events, ie, specific events accompanying system startup, include power-on, hardware reset,
Alternatively, reception of a reset command from the host or the like can be applied.

The invention relating to the above-mentioned device (disk storage device) focuses on the power management function of the device.
The threshold for power management can be considered as an invention relating to a method of performing power management (power management method) by switching and setting the threshold immediately after the occurrence of a specific event accompanying system startup and after a certain time has elapsed. is there.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment in which the present invention is applied to a magnetic disk drive will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of a magnetic disk drive according to one embodiment of the present invention. In the magnetic disk drive (HDD) of FIG. 1, reference numeral 11 denotes a disk (magnetic disk) as a recording medium on which data is magnetically recorded;
Reference numeral 12 denotes data writing (data recording) on the disk 11.
And reading data from the disk 11 (data reproduction)
(Magnetic head). Head 12
Are provided corresponding to each recording surface of the disk 11, respectively.

On the recording surface of the disk 11, a number of concentric tracks (not shown) are formed. In each track, servo areas (not shown) in which servo information used for seeking / positioning of the head 12 is recorded are arranged at equal intervals. Sectors (data sectors) as a plurality of recording units are arranged between the servo areas. The servo areas are radially arranged on the disk 11 radially across the tracks from the center.

The disk 11 is a spindle motor (hereinafter, referred to as a spindle motor).
(Referred to as SPM) 13. Head 1
Numeral 2 is attached to a head actuator (rotary head actuator) 15 as a head moving mechanism, and moves in the radial direction of the disk 11 according to the rotation (angular rotation) of the actuator 15. This allows
The head 12 is sought and positioned on a target track. The head actuator 15 has a voice coil motor (hereinafter, referred to as VCM) 14 which is a driving source of the actuator 15.
Driven by the CM 14.

On the outer peripheral side of the disk 11, (SPM 13
A ramp (not shown) for retracting the head 12 when the rotation of the disk 11 is stopped (according to the rotation stop of the disk 11) is arranged.

The SPM 13 is driven by an operation current (SPM current) supplied from an SPM driver (SPM drive circuit) 16. The head actuator 15 having the VCM 14 is driven by an operation current (VCM current) supplied from a VCM driver (head actuator drive circuit) 17. SPM driver 16 to SPM 13
The value (operation amount) for determining the operation current supplied from the VCM driver 17 to the VCM 14 is C
Determined by the PU 25.

The head 12 seeks on the target track.
After the positioning, the disk 11 is rotated to scan the track. The head 12 sequentially reads the servo information of the servo areas arranged at regular intervals thereon by scanning. The head 12 reads and writes data from and to a target sector by scanning.

The head 12 is connected to a head amplifier circuit (head IC) 19 mounted on, for example, a flexible printed wiring board (FPC). Head amplifier circuit 19
Controls switching of the head 12 (under the control of the CPU 25), input / output of read / write signals to / from the head 12, and the like. The head amplifier circuit 19 amplifies an analog output (read signal of the head 12) read by the head 12, and a R / W circuit (read / write IC) 20
A predetermined signal processing is applied to the write data sent from the controller 12 and sent to the head 12.

An R / W (read / write) circuit 20 is an AGC for amplifying an analog output (read signal of the head 12) read from the disk 11 by the head 12 and amplified by the head amplifier circuit 19 to a constant voltage. (Automatic gain control) function and a decode function (read channel) for performing signal processing necessary for decoding, for example, NRZ code data from a read signal amplified by the AGC function
And an encoding function (write channel) for performing signal processing necessary for recording data on the disk 11, and pulsing the read signal to enable extraction of servo information from the read signal and outputting it as pulsed read data And a function of extracting burst data in the servo information according to a timing signal (burst timing signal) from the servo processing circuit 21 described below. This burst data is sent to the CPU 25,
It is used for positioning control for positioning the head 12 at a target position on a target track.

The servo processing circuit 21 has a function of generating various timing signals including a burst timing signal for obtaining servo information from a read pulse output from the R / W circuit 20, and a function of generating a cylinder code in the servo information. It has a function of extracting. This cylinder code is
The data is sent to the CPU 25 and used for seek control for moving the head 12 to a target track.

The disk controller 22 is connected to a host system (hereinafter, referred to as a host) such as a PC using an HDD. The disk controller 22
An interface control function for controlling communication of commands (write commands, read commands, etc.) and data with the host, a disk control function for controlling data transfer to and from the disk 11, and a buffer memory 23 described below.
And a buffer control function for controlling the

The buffer memory 23 mainly includes a write cache for temporarily storing data (write data) transferred from the host and written to the disk 11 and data (read) read from the disk 11 and transferred to the host. Data) is used as a read cache for temporarily storing data. The buffer memory 23 is configured using, for example, a RAM (Random Access Memory).

The CPU 25 has an HD according to the control program.
D overall control, for example, seek / positioning control of the head 12 based on the cylinder code extracted by the servo processing circuit 21 and burst data extracted by the R / W circuit 20, and by the disk controller 22 according to a read / write command from the host It executes disk access control (read / write access control), load / unload control of the head 12, and the like.

The power management in the present embodiment, in which the CPU 25 performs power management directly related to the present invention, is performed in three stages as in the prior art. However, in this embodiment, immediately after the occurrence of a specific event that accompanies system startup such as power-on,
This is different from the conventional method in that the power management operation threshold is switched after a certain time has elapsed.

The target of power management in the above three stages is R / W
(Read / Write) Circuit System 31, Positioning Circuit System 3
2 and the SPM drive circuit system 33. R / W circuit system 3
1 includes an R / W circuit 20, a disk control function unit and a buffer control function unit of a disk controller 22, and a buffer memory 23. The positioning circuit system 32 includes a VCM
It includes a driver 17, a head amplifier circuit 19, and a servo processing circuit 21. This positioning circuit system 32 includes R /
Part of the W circuit 20 (for example, a burst data extraction function unit)
Is also included. The SPM drive circuit system 33 includes the SPM driver 16. A power supply line to each of the circuit systems 31 to 33 has a power management mechanism together with a power management function of the CPU 25, and a changeover switch (not shown) such as an FET switch (controllable by the CPU 25).
Is provided. ON / OFF of this switch realizes connection / disconnection of (power supply lines to) the circuit systems 31 to 33.

The CPU 25 includes a flash ROM (Read Only Memory) 26 as a rewritable nonvolatile memory in which the control program is stored in advance, and a RAM (Random A) for providing a work area of the CPU 25 and the like.
ccess Memory) 27 is connected.

Next, the magnetic disk drive having the configuration shown in FIG.
The flow of the entire power management process in FIG.
This will be described with reference to a chart. First, power on (O
N) or a hardware reset (for
Switch operation), or a reset instruction is issued from the host.
(For example, when the reset command is received)
A specific event involving the launch of the system (hereinafter, this specific event
Is conveniently turned on, that is, turned on.)
Occurs, immediately after the power is turned on (the system
Management time t optimized for operation at startup) ₁Works
A threshold is set (step S1).

[0041] Next CPU25 proceeds to step S2, if there is Easy going request from the host such as a personal computer running the service, to operate the power management mechanism when the access request is interrupted threshold t ₁ or more currently configured .

The details of steps S1 and S2 will be described later.

If there is no access request and the power management mechanism is not operated in step S2, the CPU 25
Proceeds to step S3, a certain time from power t _p
Check if has passed. If the time t _p
If has not elapsed, the process returns to step S2, and if it has elapsed, the process proceeds to step S4.

The CPU25 is proceeds to step S4, the operation state after the time t _p has elapsed from when the power is turned on, i.e., optimized management time t ₂ the operating threshold of the normal operating state after system startup (normal state) Set as Here, it is general that t ₁ and t ₂ satisfy t ₁ > t ₂ .

Next, the CPU 25 proceeds to step S5, executes the service if there is an access request from the host, and executes the service if the access request is interrupted by the currently set threshold value t ₂ or more, as in step S2. Activate

If there is no access request and the power management mechanism is not operated in step S5, the CPU 25
Proceeds to step S6, and checks whether the power is turned off (OFF). If the power has not been turned off, the process returns to the step S5, and if the power has been turned off, a series of processing ends.

[0047] To summarize the overall flow of the above power management, the power management mechanism, the operating threshold t ₁ to be applied from power during the predetermined time t _p, the operation threshold t ₂ applied to it after And two thresholds. here,
If the t ₁ larger than t _p, never power management mechanism is actuated at least until the time t _p has elapsed.

Next, details of steps S1 and S2 will be described with reference to the flowchart of FIG. In this flowchart, the power management by the power management mechanism is performed in three stages as in the related art, that is, the first stage involves disconnection of the R / W circuit system 31, and the second stage involves (including the retraction of the head 12). The third stage of disconnecting the positioning circuit system 32 is disconnecting the SPM drive circuit system 33, that is, stopping the rotation of the SPM 13. The return at the first, second, and third stages requires a microsecond (μs) unit, a millisecond (ms) unit, and a second (s) unit, respectively.
Therefore, the power management mechanism needs an optimum threshold value for each of the first, second, and third steps.

The step S1 comprises steps S11 and S12 as shown in FIG. CPU25
, In step S11, sets the operation threshold t ₁ for operating a power management mechanism. As described above, the first, second, because it is necessary each optimum threshold corresponding to the third stages of, t ₁ is actually three numbers base vector value (t _11, t ₁₂ , T ₁₃ ). Here, t ₁₁ <t ₁₂ <t
It is ₁₃ .

The CPU 25 determines in step S11 that the threshold t
_{When 1} (t ₁₁ , t ₁₂ , t ₁₃ ) is set, step S 12
Then, the timer value of a timer (not shown) is set to 0 and the timer is started. Then, the timer starts counting time. The timer value of this timer indicates the elapsed time from the time when it was set to 0.

The CPU 25 executes step S12, that is, step S1 consisting of steps S11 and S12.
Is executed, the process proceeds to step S2. This step S2
Comprises steps S13 to S27 as shown in FIG.

First, the CPU 25 checks whether there is an access request from the host (step S13).
If there is an access request, the flow branches to step S14; otherwise, the flow branches to step S22.

In step S14, the CPU 25 checks whether or not the SPM drive circuit system 33 is disconnected and the SPM 13 is stopped in a step S25 described later. If the SPM 13 is stopped, the process proceeds to step S17.
If not, the flow branches to step S15.

In step S15, the CPU 25 checks whether or not the positioning circuit system 32 has been disconnected in step S26 described later.
If not, the flow branches to step S16.

In step S16, the CPU 25 checks whether or not the R / W circuit system 31 has been disconnected in step S27 to be described later.
If not, the flow branches to step S20.

When an access request is made and the power management mechanism is already operating and in the power saving state, if the power saving state is in the third stage, that is, the R / W circuit system 31, Positioning circuit system 32 and SPM drive circuit system 33
Are all disconnected, the CPU 25 first drives the SPM drive circuit system 33 in step S17 to
Then, in step S18, the positioning circuit system 32 is activated to enable the seek / position control of the head 12, and in step 19, the R / W circuit system 3 is turned on.
1 is enabled so that the disk 11 can be read / written.

Similarly, if the power saving state is in the second stage, that is, the R / W circuit system 31 and the positioning circuit system 32
Are disconnected, the activation of the positioning circuit system 32 in step S18 and the activation of the R / W circuit system 31 in step S19 are sequentially performed. If the power saving state is in the first stage, that is, if only the R / W circuit system 31 is disconnected, only the activation of the R / W circuit system 31 in step S19 is performed.

When step S19 is executed, the process proceeds to step S20, similarly to the case where it is determined in step S16 that the R / W circuit system 31 has not been disconnected. In step S20, the CPU 25 executes processing for an access request from the host. When the processing for the access request is completed, the CPU 25 clears the timer to 0 and restarts (step S21), and thereafter returns to step S13.

When there is an access request from the host, the execution of steps S17, S18 and S19 delays the access processing. Therefore, it is clear from the flowchart of FIG. 3 that the balance between power saving and access efficiency is important.

On the other hand, at step S13, the host
If it is determined that there is no access request, the CPU 25
Ima value and threshold t₁₃Is compared with the timer value.
Is t ₁₃It is checked whether it is greater than (step S2
2). Here, the timer starts counting from the time when the initial value t1 is set.
Or the end of the last access processing (in step S20)
The elapsed time from the time is shown.

[0061] If the timer value is to step S25 if t ₁₃ larger, the process branches to step S23 if t ₁₃ below.

[0062] CPU25 at step S23, performs comparison between the timer value and the threshold value t _12, the timer value is checked whether t ₁₂ greater than the timer value t ₁₂
If greater than the step S26, the process branches to step S24 if t ₁₂ below.

[0063] CPU25 in step S24, performs comparison between the timer value and the threshold value t _11, the timer value is checked whether greater than t _11, the timer value t ₁₁
If greater than the step S27, and ends the if t ₁₁ following the whole step S2 processing (processing and power management mechanism for the access request).

[0064] Now, if the timer value is greater than t ₁₃ (step S22), CPU 25 disconnects the SPM driving circuit system 33 actuates the power management mechanism in step S25,
The rotation of the SPM 13 is stopped, and thereafter, step S2
The whole process ends.

[0065] Further, when the timer value is (t ₁₃ in the following) t ₁₂ is larger than (step S23), CPU 25 is (after retracting the head 12 to a predetermined position of the lamp) at step S26 actuates the power management mechanism Then, the positioning circuit system 32 is disconnected, and thereafter, the entire process of step S2 ends.

[0066] In addition, (at t ₁₂ below) timer value when t ₁₁ is larger than (step S24), CPU 25 disconnects the R / W circuitry 31 actuates the power management mechanism in step S27, step thereafter The whole process of S2 ends.

In the flowchart shown in FIG. 3, it is implicitly assumed that the time at which step S2 (the middle step S13) is performed is t ₁₁ or less.
In this case, (timer value is determined to be greater than t ₁₃₎
When step S25 is executed, step S2 is already executed.
7, S26 has also been executed. If this is not guaranteed, when control is transferred to step S26,
It is checked whether or not the disconnection of the R / W circuit system 31 has already been executed. If the disconnection has not been executed, a process that proceeds to step S26 after the execution may be added. Step S25
The same applies when the control is shifted to the above. It is sufficient to check whether or not the disconnection of the positioning circuit system 32 has already been executed, and if it has not been executed, add a process to proceed to step S25 after the execution.

Note that the details of steps S4 and S5 in the flowchart of FIG. 2 are the same as the details of steps S1 and S2, and a description thereof will be omitted. If necessary, the read as respectively S4, S5 in the flowchart S1, S2 of FIG. _{3, t 1, t 11,} t 12, t
_By replacing ₁₃ with t ₂ , t ₂₁ , t ₂₂ , and t ₂₃ , the flowchart of FIG. 3 can be treated as a detailed flowchart of steps S4 and S5.

The power management in the magnetic disk drive having the configuration shown in FIG. 1 has been described above. As described in the section of [Problems to be Solved by the Invention], the frequency of access from the host to the magnetic disk device differs depending on the usage state.
On the other hand, during a system start-up period when the host side is turned on, the access frequency distribution is determined regardless of the use state.

[0070] Thus, as in the present embodiment, the period of predetermined time t _p from power, that after the said predetermined time t _p and setting switching threshold of the power management mechanism for,
Optimal power saving management can always be realized without causing unnecessary operation delay in response to an access request from the host.

[Modification] The optimum threshold value depends on the use state of the magnetic disk device, for example, the access frequency distribution (here, the access interval frequency distribution, that is, the frequency distribution of the time interval at which an access request is sent from the host). Therefore, the value may change every moment.

Therefore, if the optimum threshold value is calculated from the latest access frequency distribution (access interval frequency distribution) of the magnetic disk device and the threshold value is dynamically changed sequentially, effective power management can be realized. It becomes possible.

However, with this method alone, for example, when the power is turned on, if the threshold dynamically determined immediately before the power is turned off is used as it is, the access frequency differs between when the system is started and after the system is started. , Cannot be the optimal threshold.

Therefore, after starting the system,
After power predetermined time t _p has elapsed from the time dynamically changes the threshold based on the latest access frequency (access interval frequency distribution), earlier the previous embodiment and similarly fixed threshold t ₁ ( A modified example of the above embodiment using (t ₁₁ , t ₁₂ , t ₁₃ ) will be described with reference to the flowchart of FIG.

First, at the time of power-on (ON) (that is, at the time of the original power-on, at the time of hardware reset, or at the time when a specific event involving system startup occurs, such as when a reset command is issued from the host). 2 are the same as steps S1, S2, and S3 in the flowchart of FIG.

Here, the management time t ₁ optimized for the operation immediately after the power is turned on is set as the operation threshold, and thereafter,
Until time t _p has elapsed, if any Easy going request from the host to execute the service, the process of operating a power management mechanism when the access request is interrupted threshold t ₁ or more is performed. However in the case of t _1> t _p, as described in the previous embodiments, from power-on time to time t _p has elapsed, it does not actuate the power management mechanism.

[0077] Now, the time t _p from power point has elapsed (step S33), CPU 25 sets the threshold value t ₂ (step S34). This t ₂ may be either a predetermined value as in step S4 in FIG. 2 or a dynamic optimum value calculated immediately before the power is turned off prior to the current power-on.

Next, the CPU 25 proceeds to step S2 in FIG.
Similarly, if there is Easy going request from the host executes the service, to operate the power management mechanism when the access request is interrupted threshold t ₂ or more (step S35).

On the other hand, when there is no access request and the power management mechanism is not operated, the CPU 25
_It is determined whether or not the recalculation time ₂ has arrived, for example, based on whether or not the number of accesses has exceeded a predetermined number of times since the previous change of the threshold (step S36). It should be noted that other criterion, for example, a simple criterion for determining whether or not the elapsed time since the previous change of the threshold value has exceeded a predetermined time may be used.

[0080] If the if recalculation timing, CPU 25 recalculates the optimal management time to the current access interval frequency distribution (step S37), the t ₂ is updated to a new value the recalculated (step S38 ).

Here, the optimum management time (optimum value) is defined as the minimum time during which the processing speed is not reduced by a predetermined N% or more from the access interval frequency distribution (within a limit not exceeding a predetermined range). Used. Details of this (optimum value) determination will be described below.

First, an example of the access interval is shown in FIG.
As shown in the figure, t _a , t _b , t _c ... Are access intervals, and are time intervals of access requests (for example, read / write commands) issued from the host to the magnetic disk device of FIG.

FIG. 6 shows an example of the access interval frequency distribution.
Shown in As shown in the figure, the access interval frequency distribution is
At predetermined intervals, an access interval of an access request issued from the host during the predetermined period is classified into i times (i = 1, 2, 3,...) Times unit time t ₀ and it _0. Te, is obtained by taking the number of occurrences n _i of each time it ₀ another access interval. As can be seen, the power management mechanism is set to the threshold t
_When operated at ₂ , t ₂ is kt ₀ ≦ t ₂ <(k + 1)
If it is assumed that t _0, the access interval is kt ₀ or more access request is executed with a delay of recovery time t _r of the power management mechanism.

[0084] In this case, the sum of the access interval of all access requests access interval frequency distribution of FIG. 6 and T1, the access interval is kt ₀ or more of the total access request recovery time (loss time) the sum of t _r and T2 Then, the reduction rate (processing loss) L of the processing when the power management mechanism is operated at the threshold value t _2.
(%) Is represented by the following equation: L = (T2 / T1) × 100 (%) (1)

In step S37, the threshold value t ₂ is varied on the access interval frequency distribution, and the minimum time within the range where the value L does not exceed the predetermined N% is set to the optimal value for the current access interval frequency distribution. Is calculated as The optimum value is set as a new administrative time t ₂ in step S38. Note that t ₂ is 3t ₀ ≦ t ₂ <4t ₀ as in the example of FIG.
, T1 is n ₁ × t ₀ + n ₂ × 2t ₀ + n ₃ × 3t ₀ + n
₄ × 4t ₀ + n ₅ × 5t ₀ +..., And T2 is (n ₄ + n ₅
+ ...) it is represented by t _r.

In addition, N satisfying N1 <N <N2
1, N2 is set, and the above-mentioned L is calculated by applying the current threshold value to the latest access interval frequency distribution. If the value of L is equal to or more than N2 (that is, if the processing efficiency is relatively low) The current threshold value is updated to a value increased by a first predetermined value (first predetermined time) (the processing efficiency is increased from the current value), and if the value of L is equal to or less than N1, the processing efficiency is relatively low. If it is higher, the current threshold may be updated to a value reduced by a second predetermined value (a second predetermined time) (thus lowering the processing efficiency from the present). Note that the first and second predetermined values may be the same value or different values.

When the CPU 25 updates t ₂ to the optimum value calculated in step S37 (step S38), the process returns to step S34 and sets t ₂ as a new threshold.

On the other hand, if the recalculation time has not arrived, the CPU 25 checks whether or not the power has been turned off (step S39). If not, the process returns to step S35. The process ends.

[0089] As described above, in this modified example, after a time t _p has elapsed from the time of power on (i.e. during normal use) is optimum threshold from the access interval frequency distribution (i.e. use state of the magnetic disk apparatus of FIG. 1) while achieve effective power management by dynamically changing the calculated and sequentially threshold, immediately after power-on (period from the time when the power is turned t _p) were dynamically determined before power up Effective power management is realized by using a threshold independent of the threshold. As described above, in this modified example, at the time of power-on (at the time of original power-on, at the time of hardware reset, or at the time of occurrence of a specific event involving system startup such as when a reset command is issued from the host), More optimal power management can be realized both when the system is started upon power-on and during normal use after the system is started.

The calculated optimum value t ₂ is set so that the optimum value t ₂ does not disappear when the power is turned off.
For example, a predetermined area called a system area that cannot be accessed by a user, or a rewritable nonvolatile memory,
For example, it is stored in a predetermined area of a flash ROM (FROM) 26.

In the above-described embodiment, the case where the threshold value of the power management mechanism is determined on the magnetic disk device side has been described. However, the present invention is not limited to this. Can also be specified. That is, since the above threshold value depends on the processing speed of the host, the processing speed of the magnetic disk device, and the usage pattern, the host should specify a value obtained at the time of design evaluation when the system is started up. Is possible. However, also in this case, it is not preferable to use the threshold specified by the host as it is when starting up the system.

[0092] Therefore, if the threshold from the host is specified, it will only change the t ₂ without threshold changing if during the execution of step S2 in FIG. 2, if the time of execution of step S5 t if the threshold value is also changed to the t ₂ while changing the _2, effective power management, including the startup of the system can be realized.

Although the present invention has been described with reference to the case where the present invention is applied to a magnetic disk device, the present invention relates to a disk storage device other than a magnetic disk device such as a floppy (registered trademark) disk device, a CD-ROM device, and a magneto-optical disk device. The same applies to devices.

[0094]

As described above in detail, according to the present invention, the operation threshold value for power management is switched between immediately after the occurrence of a specific event such as power-on and the like and at the time after a certain time has elapsed. Therefore, optimal power saving management can always be realized while minimizing the operation delay for an access request from the host.

According to the present invention, the operation threshold value for power management is dynamically changed and set based on the use state of the apparatus after the lapse of the predetermined time, so that the operation threshold value in the normal state after system startup is established. More optimal power saving management can be realized while suppressing a decrease in performance.

Further, according to the present invention, for a certain period immediately after the occurrence of the specific event, a threshold determined independently of the threshold dynamically changed and set is used. More effective power saving management can be performed without sacrificing system startup time.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a magnetic disk drive according to an embodiment of the present invention.

FIG. 2 is an exemplary flowchart for explaining the overall processing flow of power management in the embodiment.

FIG. 3 is a flowchart for explaining details of steps S1 and S2 in FIG. 2;

FIG. 4 is a flowchart for explaining a modification of the embodiment.

FIG. 5 is a diagram illustrating an access interval.

FIG. 6 is a diagram for explaining an access interval frequency distribution.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 11 ... Disk 12 ... Head 13 ... SPM (spindle motor) 16 ... SPM driver 17 ... VCM driver 19 ... Head amplifier circuit 20 ... R / W circuit (read / write circuit) 21 ... Servo processing circuit 22 ... Disk controller 23 ... Buffer Memory 25 CPU 31 R / W circuit system 32 Positioning circuit system 33 SPM drive circuit system

Claims (10)

[Claims] When an access request from a host is interrupted for more than a time specified by a power management threshold, a predetermined portion in the device is set to a power saving state, and access from the host is performed in the power saving state. Power management means for returning to the original state when a request is generated; and threshold switching setting means for switching and setting the power management threshold immediately after the occurrence of a specific event accompanied by system startup and after a lapse of a certain period of time. A disk storage device comprising: 2. When an access request from a host is interrupted for more than a time specified by a power management threshold, a predetermined portion in the device is set to a power saving state, and access from the host is performed in the power saving state. Power management means for returning to the original state when a request is generated; and threshold switching setting means for switching and setting the threshold for power management immediately after the occurrence of a specific event involving system startup and after a certain period of time has elapsed. And a threshold switching setting means for dynamically changing and setting the threshold after the lapse of the predetermined time according to the use state of the disk storage. 3. The apparatus according to claim 1, further comprising a suppression unit for preventing the power management unit from setting a predetermined portion in the device in a power saving state for the predetermined time immediately after the occurrence of the specific event. The disk storage device according to claim 1 or 2. 4. The threshold switching setting unit includes the inhibiting unit, and the inhibiting unit sets the threshold for power management immediately after the occurrence of the specific event to a value greater than the predetermined time. 4. The disk storage device according to claim 3, wherein the predetermined portion in the device is prevented from being set to a power saving state by the power management means for the predetermined time immediately after the occurrence of the specific event. 5. The threshold switching setting means uses a threshold determined independently of the threshold dynamically changed and set for the certain period of time immediately after the occurrence of the specific event. 3. The disk storage device according to 2. 6. When an access request from a host is interrupted for more than a time specified by a power management threshold, a predetermined portion in the device is set to a power saving state, and access from the host is performed in the power saving state. Power management means for returning to the original state when a request is generated; and threshold switching setting means for switching and setting the power management threshold immediately after the occurrence of a specific event accompanied by system startup and after a lapse of a certain period of time. A certain time immediately after the occurrence of the specific event is set as a threshold value for the power management to a value larger than the certain time, or a threshold value determined independently of a threshold value specified from the outside And threshold switching setting means for setting a threshold specified externally after the predetermined time has elapsed. Disk storage device according to claim. 7. The disk storage device according to claim 1, wherein the specific event is power-on, hardware reset, or reception of a reset command from the host. 8. When an access request from a host is interrupted for a time longer than a time specified by a threshold for power management, a predetermined portion in the apparatus is set to a power saving state, and access from the host is performed in the power saving state. A power management method for a disk storage device that returns to an original state when a request occurs, wherein the power management threshold value is switched between immediately after a specific event accompanied by system startup and after a lapse of a certain period of time. A power management method characterized by setting. 9. The power management method according to claim 8, wherein the threshold value is dynamically changed and set according to a use state of the disk storage device after the predetermined time has elapsed. 10. The power management method according to claim 9, wherein a threshold determined independently of the threshold dynamically changed and set is used for the predetermined time immediately after the occurrence of the specific event.