JP3068811B2 - Method of controlling storage device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a control technique for a storage device, and more particularly to a technique effective for speeding up access to a plurality of pieces of information at distant positions in the same information unit.

[0002]

2. Description of the Related Art For example, in a general-purpose computer system, it is known to use a large magnetic disk device as an external storage device.

In such a magnetic disk drive, a plurality of tracks are provided concentrically on each of a plurality of coaxially fixed and rotating magnetic disks, and each magnetic disk is equidistant from the center of rotation. In a cylinder unit, which is a logical concept formed by connecting a plurality of tracks that can be accessed at relatively high speed only by a head selection operation without performing a seek operation by moving the head in the radial direction of the magnetic disk, It is common to manage data.

As a data recording format for each track, a so-called CKD system is mainly used.

That is, each of a plurality of records stored in a track is constituted by a count section C, a key section K and a data section D arranged at a predetermined gap, and the count section C stores the record. In addition to recording positioning information in the apparatus and management information such as the length of the key portion K and the data portion D, index information for searching the data portion D is recorded in the key portion K, and actual data is recorded in the data portion D. Is stored.

[0006] While the head passes through each gap, preparation for a predetermined process is continuously performed on a part that comes next.

[0007] Recently, as disk drives have become more fixed and larger in capacity, there has been a strong demand for high-speed data transfer from customers, and in response to this, a plurality of tracks are collectively processed (continuous processing). While trying to prevent rotation waiting by processing tracks one track at a time, a dump operation to save data from a magnetic disk to a magnetic tape or the like, or a transfer operation from a magnetic tape to an original magnetic disk or from a magnetic disk to a magnetic disk. In restore processing for restoring data, by selectively dumping / restoring only the updated tracks instead of processing all tracks in its entirety, the processing time is shortened and the number of tape windings is reduced by reducing the amount of data. Tend to be.

For this purpose, it is necessary to collectively process discontinuous tracks on the same cylinder.

For this reason, in the magnetic disk drive adopting the CKD method as described above, tracks are switched within the same cylinder, for example, SEEK H
An EAD command or a command with a multi-track instruction is used.

[0010]

However, the former SE
Since the EK HEAD command needs to receive the individual track number from the upper central processing unit every time it is executed, it cannot be processed in the gap unique to the CKD method, and the read / write head is repositioned by a subsequent read command. It needs to be corrected, and it is impossible to process the next record without waiting for rotation. And this SEEK
The HEAD command needs to be performed before the read command, and there is a problem that the overhead cannot be ignored.

On the other hand, the latter command with a multi-track instruction performs an operation of switching the head at the track start point (index marker) while guaranteeing the position of the read / write head, but always selects the next adjacent track. Therefore, in order to continue processing for a track distant from the current track, the command with the multi-track instruction is issued a plurality of times, and the number of times of issuing the command with the multi-track instruction until the index is reached in order to switch tracks. There is a greater problem that the more the distance between the current track and the next target track, the more the rotation wait occurs.

Accordingly, conventionally, the former SEEK HE
Although the AD command is used, this command has a large overhead as described above. To absorb this overhead at the time of track switching, a large gap is required behind the track, and repositioning of the head is also required.

This is because, in recent years, when a data format that uses almost the entire track is adopted in order to secure the storage capacity of the magnetic disk drive, one time is always required to continuously process non-adjacent tracks. It is inevitable that a wasteful rotation time occurs, which causes an increase in the time required for data transfer between the central processing unit and the magnetic disk device.

As a data recording method for avoiding rotation waiting in a magnetic disk device, for example,
There is a technique disclosed in Japanese Patent Application Laid-Open No. 83967.

That is, in this technique, a position of an index marker of a track is set to be shifted in each of a plurality of cylinders, so that a track of a certain cylinder is
The object of the present invention is to avoid the rotation waiting that occurs during the operation of moving and switching the head to the track of the next cylinder.

However, according to this technique, since the positions of the index markers of the respective tracks in the individual cylinders are the same, there is no effect in avoiding the rotation waiting in the head switching in the same cylinder.

SUMMARY OF THE INVENTION An object of the present invention is to provide a storage device control technique that enables continuous access to a plurality of pieces of information that are separated from each other in the same information unit. Another object of the present invention is to provide a control technology of a storage device capable of high-speed access when executing a read command for accessing a plurality of discontinuous positions on a disk device.

The above and other objects and novel features of the present invention will become apparent from the description of the present specification and the accompanying drawings.

[0019]

The following is a brief description of an outline of a typical invention among the inventions disclosed in the present application.

That is, the present invention is a method for controlling a storage device having a disk device for storing information and a disk control device connected to the higher-level processing device for controlling the disk device. Receiving a command specifying a plurality of discontinuous pieces of information on the disk device before receiving the command, and transferring the information on the specified disk device from the disk device to the disk control device. It is. The present invention also relates to a method for controlling a storage device having a disk device for storing information and a disk control device connected to a higher-level processing device for controlling the disk device, wherein a read command in CKD format is transmitted from the higher-level processing device. Receiving a command including a parameter for specifying a plurality of pieces of information on the disk device before receiving the command, and transferring information on the specified disk device from the disk device to the disk control device. It is. The present invention is also a method of controlling a storage device having a disk device for storing information and a disk control device connected to a higher-level processing device for controlling the disk device, wherein the method receives a read command from the higher-level processing device. Receiving a command including a parameter specifying a plurality of pieces of information on the disk device, and transferring the information on the disk device specified by the parameter from the disk device to the disk control device. Control method. Further, the present invention provides a control method of a storage device having a disk device for storing information, a head for accessing the stored information, and a disk control device connected to a host processing device for controlling the disk device and the head. Receiving a second command including a parameter specifying a plurality of pieces of information on a disk device before receiving a first command in a CKD format for reading information in a storage device from a host processing device; Moving the head to a position on the disk device for the information specified by the second command; and transmitting the information from the storage device to the host processor in response to the first command. . The present invention also provides a disk device for storing information,
A method for controlling a storage device having a head for accessing information and a disk controller connected to a higher-level processing device for controlling the disk device and the head, wherein the disk device is controlled before a read command is received from the higher-level processing device. Receiving a command specifying a plurality of discontinuous pieces of information above, and moving a position of a head to access information on the specified disk device before receiving a read command from a higher-level processing device And The present invention also relates to a method for controlling a storage device having a disk, a head for accessing information stored on the disk, and a disk control device for controlling the disk and the head, the method comprising receiving a read command from a host processor. Receiving a command specifying the position of a plurality of discontinuous pieces of information on the disk, controlling the head to the position of the specified information, and sending a command from the disk control unit to the host processing unit according to the read command. Transmitting the specified information. The present invention also relates to a method for controlling a storage device having a disk, a head for accessing information stored on the disk, and a disk control device for controlling the disk and the head, the method comprising receiving a read command from a host processor. Receiving a command including a parameter specifying a plurality of pieces of information on the disk, controlling a head to read the specified information, and specifying a command from the disk control device to the higher-level processing device according to the read command. Transmitting the information.

According to the storage device control method of the present invention described above, the disk controller can know the positions of a plurality of discontinuous pieces of information before receiving a read command. As a result, a high-speed access to a plurality of pieces of information becomes possible at the time of executing the actual read command. In addition, since the position of the information of the reading destination is known before the head position for reading is specified, the control of the head is quickened.

[0022]

Embodiments of the present invention will be described below in detail with reference to the drawings.

(Embodiment) FIG. 1 is a block diagram showing an example of a system configuration when a storage control system for implementing a storage device control method according to an embodiment of the present invention is incorporated in a general-purpose computer system. FIG. 2 is a flowchart showing an example of the operation.

For example, a central processing unit 1 of a general-purpose computer has a disk control unit 3 and a magnetic disk unit via a channel unit 2 for controlling transmission and reception of data and commands between the central processing unit 1 and the outside. 4 (rotary storage device).

The control of data input / output between the central processing unit 1 and the magnetic disk unit 4 via the channel unit 2 and the disk control unit 3 is controlled by the central processing unit 1 via the channel unit 2. This is performed by the CCW chain 50 composed of a series of channel command words (hereinafter simply referred to as CCWs or commands) issued to the disk controller 3.

That is, the disk control device 3 includes a command decoding unit 31 for analyzing individual commands coming from the channel device 2, a track switching unit 32 activated by a control signal 31 a from the command decoding unit 31,
Similarly, a disk read / write processing unit 33, which is started by a control signal 31b from the command decoding unit 31 and controls data transfer between the magnetic disk device 4 and the channel device 2, is provided.

The disk control device 3 is provided with a track switching map 34 having a predetermined bit width.

In this embodiment, for the sake of simplicity of description, the components of the disk control unit 3 are divided into a command decoding unit 31, a track switching unit 32, a disk read / write processing unit 33, a track switching map 34 and the like. Although separately illustrated, these functions can be configured by, for example, a general-purpose or dedicated microprocessor, and a program and a memory that control the microprocessor.

In this case, the track switching map 34 is configured to hold bit information 34a corresponding to the number of tracks for one cylinder in the magnetic disk drive 4.

That is, the magnetic disk device 4 in this embodiment is provided with eight magnetic disks (rotary storage media) (not shown) whose both surfaces are recording surfaces, and records servo information for controlling the positioning of the head. 15 excluding the burned surface
15 equidistant from the center of rotation on each of the recording surfaces
, Each track is provided with a track number from # 0 to #E, and bit information 34a corresponding to each of them is held in the track switching map 34. .

In the track switching map 34, a DEFINE TRACK SELECT which will be described later is used.
When the control signal 31c sent from the command decoding unit 31 when the T command 51 arrives, the DEF
An operation is performed in which the bit information 34a of the track number #n corresponding to a plurality of tracks to be accessed, designated as parameters accompanying the INE TRACK SELECT command 51, is set to "1" and the others are set to "0". .

The track switching map 34 is connected to the track switching section 32 via a signal line 35.
Operation for instructing the magnetic disk device 4 to switch a head (not shown) to the magnetic disk device 4 without performing a redundant operation of receiving a track number from the higher-order channel device 2 every time the track is switched based on the bit information 34a in 4. Is what you do.

FIG. 3 is a diagram showing an example of the CCW chain 50 used in the control method of the storage device of the present embodiment.

First, DEFINE TRACK SEL
The ECT command 51 is a command peculiar to the present embodiment.
As described above, the bit information corresponding to the track number #n corresponding to the track switching map 34 provided in the disk control device 3 is held by setting the bit information to "1".

The SEEK command 52 is a command for performing a seek operation for positioning the head to a target cylinder by moving the head in the radial direction of a magnetic disk as a storage medium.

Further, the SEARCH ID command 53
Is a command for positioning the head at the target record in the track. That is, CK as in the present embodiment is used.
In the recording method based on the D method, the operation of reading the count portion of each record and checking whether or not the record is a target record based on identification information such as a cylinder number, a track number, and a record number recorded in the count portion is performed. Is what you do.

The TIC command 54 performs a conditional branch operation in the CCW chain in the CCW chain 50 in accordance with the end condition of the immediately preceding command. In the case of this embodiment, the preceding SEARCH
SEARCH again when ID command 53 fails
An operation for performing the ID command 53 is performed.

Further, a READ MCKD command 55
Is a command for reading out all records from the current head position on an arbitrary track.

The MSEARCH ID command 56
Is a SEARCH ID command with a multi-track instruction.

That is, the MSEARCH ID command 56 normally detects the index marker 6a indicating the start (end) of each track 6 shown in FIG. The track switching operation is automatically performed.

An example of the operation of the storage control system and the control method according to the present embodiment will be described below.

First, from the central processing unit 1 to the channel unit 2
When the DEFINE TRACKSELECT command 51 is issued via the command, the command decoding unit 31 of the disk control device 3 decodes this command, sends out the control signal 31c, and activates the track switching map 34. This triggers the DEFINE TRACK SELE
Information on a combination of a plurality of tracks 6 specified as a parameter accompanying the CT command 51 is set in individual bit information 34 a of the track switching map 34.

Next, the head is positioned at a target cylinder in the magnetic disk drive 4 by the SEEK command 52.

Further, a SEARCH ID command 53
An operation for searching for the target record R _n in the target track 6 is performed by the TIC command 54. This search operation is repeated until the target record is found, and after the target record is found, the search operation is repeated. , Followed by REA
According to the DMCKD command 55, an operation of reading all the records in the current track up to the end of the track after the current record is performed, and the read record is transmitted to the channel device 2 via the disk read / write processing unit 33. Sent to the side.

Further, MSEARCH ID command 5
6 receives the control signal 31a.
The track switching unit 32 and the track switching map 34 are activated through the steps 31b and 31b.

Thus, the track switching section 32 refers to the bit information of the track switching map 34 and knows the track number #n of the next track to be switched.

Then, the track switching unit 32 determines which corresponding head is to be selected based on the track number #n taken from the track switching map 34.

The head number matches the track number in the cylinder. For example, in the case of this embodiment, if the track corresponding to the currently selected head is track number # 1, the track switching map 34 Since the next bit information 34a of "1" corresponds to the track number # 4, the head corresponding to the track number # 4 to be processed next from the current track number # 1 is selected.

At the same time, the track switching section 32 searches the track switching map 34 for a head (track) to be processed after the currently selected head (track).

This operation is automatically performed when the index marker 4a is detected during the execution of the MSEARCH ID command 56, regardless of whether the bit of the track number #n corresponding to the currently selected head is "1" or "0". Be executed.

At this time, DEFINE TRACK
If the SELECT command 51 has not been issued,
As before, the next track number # n + adjacent to the track with the track number #n corresponding to the currently selected head
The head of one track is selected.

On the other hand, DEFINE TRACK
If the SELECT command 51 has been issued,
The next head corresponding to the next target track number #n is obtained by referring to the track switching map 34, and the track switching unit 32 instructs the magnetic disk device 4 to select the head.

Here, the head selecting operation performed by the track switching unit 32 with reference to the track switching map 34 is performed by the conventional MSEARCH ID command 56 no matter how far the corresponding track is within the current cylinder. The operation and the processing time for selecting the head corresponding to the track adjacent to this track are the same.

Thus, the conventional MSEARCH ID
Similarly to selecting a head corresponding to a continuous track by the command 56, a head corresponding to a track at an arbitrary position distant from the current track can be selected in the same cylinder. This eliminates the need to wait for rotation each time a track is switched, for example, when continuously accessing tracks at discrete positions.

As a result, the time required to transfer data such as records between the central processing unit 1 and the magnetic disk device 4 is reduced, and the performance of the storage control system is improved.

This will be described with reference to the flowchart of FIG. 2. First, the DEFINE TRACK SELE
After the execution of the CT command 51 (step 701), the target processing is performed on the record (step 702).

Next, it is determined whether or not the command is the MSEARCH ID command 56 which is a command with a multi-track instruction (step 703).
It is checked whether the EFINE TRACK SELECT command 51 has been executed (step 704).

Then, if executed, the DE
The track switching map 34 is already set by the parameter accompanying the FINE TRACK SELECT command 51.
The track number #n (head number) to be processed next is determined with reference to the bit information 34a set in the first row, and the head corresponding to the track with the track number #n is selected (step 705).

On the other hand, the MSEARCH ID command 56
If the DEFINE TRACKSELECT command 51 has not been executed before the MSEARCH I
The operation of the D command 56 is the same as the conventional one, and the head corresponding to the adjacent track obtained by adding 1 to the current track number #n is selected (step 706).

Thereafter, the MSEARCH ID command 5
6 is executed (step 707).

Thus, the CCW chain 5 of this embodiment is
0, for example, if the current track is track number # 1, as shown in FIG.
When the index marker 4a is detected when the SEARCH ID command 56 is executed, the head corresponding to the track of the track number # 4 to be processed next is automatically selected.

As a result, in the same cylinder,
It is possible to continuously process a plurality of tracks at discontinuous discrete positions without waiting for rotation.

For this reason, for example, the magnetic disk drive 4
Even if a continuous access request is made to a group of tracks at a distant position in the medium as in the case of a differential dump of the medium, a high-speed differential processing can be performed without increasing the processing time due to the rotation wait. Dump is possible.

(Reference Example) FIG. 5 is a conceptual diagram of a track 8 showing an example of a data recording method according to a reference example of the present invention, and FIG. 6 is an example of a CCW chain constituted by a conventionally known CCW. It shows.

That is, the conventional CCW chain 9
As a method of avoiding the rotation lag that has been inevitable in the past when switching between tracks 8 apart from each other without changing 0, a small-capacity gap that does not perform any processing is provided at the rear end of the track 8. Can be considered.

However, in the method of simply providing a gap,
There is a problem that the data recording area is subjected to wrinkles, and the predetermined nominal storage capacity in one track 8 is reduced.

Therefore, in the case of this embodiment, the storage capacity in the track 8 is maintained at a predetermined nominal value by slightly increasing the bit density.

That is, in a typical apparatus currently in operation, the rotation time of the magnetic disk is 3600 rpm, so the time required for one rotation is 16.7 ms, and the time required for head switching is about 100 μs. In this example, the bit density is increased by 0.6% in order to form a gap G corresponding to a head transit time of slightly more than 100 μs.

It is to be noted that increasing the bit density by about 0.6% can be easily realized with the current technical level.

FIG. 5 is a conceptual diagram showing a case where a gap G is provided at the end of each track 8 by increasing the bit density by about 0.6%.

That is, in FIG.
The size of the gap G provided on the trailing side of the SEEK HEK after the processing for all the records R in the track 8 is completed without waiting for rotation.
The time required for selecting any other track 8 (head) by the AD command 62 is covered.

Thus, for example, in the conventional CCW chain 90 shown in FIG. 6, after an arbitrary track #n is executed by the READ MCKD 61 which processes all records R after the designated record R in one track 8, , The head position 8b is the head of the incoming gap G.

Then, the switching operation to an arbitrary track 8 by the SEEK HEAD 92 is performed in the gap G, and this switching operation ends before reaching the index marker 8a in the track 8 concerned.

That is, the head position 8c immediately after the switching operation is located before the index marker 8a.

As a result, a plurality of tracks 8 at discrete positions in the same cylinder are
By using the ARCH ID 93, continuous access can be performed without waiting for rotation.

The same applies to the write operation.

It is needless to say that the same effect can be obtained by arranging the position of the gap G immediately after the index marker 8a in the track 8, that is, at the start end of the track 8.

As described above, according to the data recording method of the present embodiment, information is stored in each track 8 in two tracks.
By increasing the bit density to be stored numerically, a gap G is added to the end or start end of each track 8 without deteriorating the nominal storage capacity of each track 8 to correspond to each of the tracks 8. Since the operation of selecting any one of the tracks 8 by switching the head is completed while the head passes the gap G, the gap G provided at the leading end or the trailing end of the track 8 For example, a conventional SE
Providing the head switching time from the current track 8 to the next track 8 arbitrarily distant from the current track 8 by the EK HEAD command, thereby causing rotation waiting for a plurality of tracks 8 separated from each other in the same cylinder. The advantage is that the access can be continuously performed without any change.

Although the invention made by the inventor has been specifically described based on the embodiments, the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the gist of the invention. Needless to say.

For example, the above embodiment and the reference example may be combined.

[0081]

The effects obtained by the representative ones of the inventions disclosed in the present application will be briefly described as follows.

That is, according to the method of controlling a storage device of the present invention, there is an effect that continuous access can be realized to a plurality of pieces of information at positions separated from each other in the storage device. Further, according to the control method of the storage device of the present invention, there is an effect that high-speed access can be performed at the time of executing a read command for accessing a plurality of discontinuous positions on the disk device.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an example of a system configuration when a storage control system in which a storage device control method according to an embodiment of the present invention is implemented is incorporated in a general-purpose computer system.

FIG. 2 is a flowchart illustrating an example of an operation of a storage control system in which a storage device control method according to an embodiment of the present invention is implemented.

FIG. 3 is a diagram illustrating an example of a CCW chain of a storage control system in which a storage device control method according to an embodiment of the present invention is performed.

FIG. 4 is a conceptual diagram of a track in which records are stored in a rotary storage device in a storage control system in which a storage device control method according to an embodiment of the present invention is implemented.

FIG. 5 is a conceptual diagram of a track in a data recording method according to a reference example of the present invention.

FIG. 6 is a diagram showing an example of a conventional CCW chain.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Central processing unit, 2 ... Channel device, 3 ... Disk control device, 31 ... Command decoding part, 31a, 31b, 31
c: control signal, 32: track switching unit, 33: disk read / write processing unit, 34: track switching map, 3
4a: bit information, 35: signal line, 4: magnetic disk device (rotary storage device), 4a: index marker, 5
0: CCW chain, 51: DEFINE TRACK
SELECT command (second command), 52... S
EEK command (first command), 53 ... SEARC
H ID command (first command), 54 ... TIC command (first command), 55 ... READ MCKD
Command (first command), 56... MSEARCH
ID command (first command), 6... Track, 6a
.. Index markers, 701 to 707 processing steps, 8 tracks, 8a index markers, 8b,
8c: head position, G: gap, 90: CCW chain, R: record.

Continued on the front page (72) Inventor Toruhisa Kashima 2880 Kozu, Kozuhara, Odawara-shi, Kanagawa Prefecture Inside the Odawara Plant, Hitachi, Ltd. References JP-A-51-113607 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G11B 20/10 G06F 3/06 302

Claims (8)

(57) [Claims] A disk device for storing information;
A disk connected to the storage device and controlling the disk device
A control method of the storage device having a read control command from the host processing device.
Specifies a plurality of discontinuous pieces of information on the disk device.
Receiving the command on the disk device.
Transferring from a disk device to the disk controller
A control method for a storage device having: 2. A disk device for storing information, comprising:
A disk connected to the storage device and controlling the disk device
A control method for a storage device having a A read command of the CKD format from the upper processing unit.
Before receiving the command, multiple information
Receiving a command that contains parameters specifying the
And The information on the specified disk device is stored in the disk device.
Transferring from a disk device to the disk controller
A control method for a storage device having: 3. A disk device for storing information, and a host device
A disk connected to the storage device and controlling the disk device
A control method for a storage device having a Before receiving a read command from the host processor
Parameter to specify multiple pieces of information on the disk device
Receiving a command containing data On the disk device specified by the parameter
Information from the disk device to the disk controller
Transferring the storage device. 4. The information from the storage device to the host processing device.
4. The method according to claim 1, further comprising the step of transmitting information.
A control method of the storage device according to any one of the above. 5. A disk device for storing information,
The head that accesses the stored information and the host processor
Connected to control the disk device and the head.
A control method for a storage device having a disk control device.
hand, Reads information in the storage device from the host processor
Before receiving the first command in CKD format,
Parameter for specifying a plurality of information on the disk device
Receiving a second command comprising: The display of information specified by the second command;
Moving the head to a position on a parking device; In response to the first command, the storage device
Sending information to a host processor.
How to control the device. 6. A disk device for storing information, and said information
Connected to the head that accesses the information and the host processor,
Disk control for controlling the disk device and the head
A method for controlling a storage device having a device, Before receiving a read command from the host processor
Specifies a plurality of discontinuous pieces of information on the disk device.
Receiving a command to Receiving the read command from the host processor
Before accessing the information on the specified disk device,
Moving the position of the head to access
A method for controlling a storage device having: 7. A disk and a disk stored on said disk
A head for accessing information, the disc and the head,
Control of a storage device having a disk controller for controlling
Your way, Before receiving a read command from the host processor,
Command to specify the location of multiple discrete information items on a disc
Receiving the command; A step of controlling the head to the position of the designated information.
And According to the read command, the disk controller
Sending the specified information to the host processor.
And a method for controlling a storage device having steps. 8. A disk and a disk stored on said disk
A head for accessing information, the disc and the head,
Control of a storage device having a disk controller for controlling
Your way, Before receiving a read command from the host processor,
Contains parameters that specify multiple pieces of information on the disc
Receiving a command; A switch for controlling the head to read the specified information.
Tep, According to the read command, the disk controller
Sending the specified information to the host processor.
And a method for controlling a storage device having steps.