KR20070115485A - Cache buffer control method and disk drive using the same - Google Patents

Abstract

A method for managing a cache buffer and a disk drive using the same are provided to determine priority for changing an internal memory block of the cache buffer based on a folder layer level of a file system applied to the disk drive. A disk(12) stores data, and the cache buffer(230) temporarily stores the data read from the disk and temporarily stores the data written to the disk. A controller(240) determines the priority for replacing file data in prior to folder data and changing the folder data of a lower level than the top level among the folder data if a condition for changing the data stored in the cache buffer is generated. The controller makes the data stored in the cache buffer replaced according to the determined priority. The priority is determined by applying an LRU(Least Recently Used) algorithm among the file data or the folder data of the same level.

Description

Cache buffer control method and disk drive using the same}

1 is a plan view of a head disk assembly of a disk drive to which the present invention is applied.

2 is an electrical circuit diagram of a disk drive to which the cache buffer management method according to the present invention is applied.

3 is a flowchart of a cache buffer management method according to the present invention.

4 illustrates a folder and file association structure for explaining a file system applied to the present invention.

The present invention relates to a cache buffer management method and apparatus, and more particularly to a cache buffer management method and apparatus for increasing the cache hit ratio in consideration of the characteristics of the file system applied to the disk drive.

In general, the difference between the data processing speed of the central processing unit and the main memory (or recording medium) in the data processing system is about 100 times or more, and the difference in processing speed is compensated by the cache buffer.

The cache buffer is a device that first reads from the main memory and stores a series of data that the CPU is expected to request next, and has a faster access speed than the main memory.

The central processing unit accesses the cache buffer first to get the desired data before accessing main memory. The higher the hit rate of this estimate, the faster the system will run.

As a general cache buffer management method, LRU (Least Recently Used) algorithm is used. The LRU algorithm is a memory management algorithm that preferentially selects and removes the most recently unused information from the information stored in the cache buffer when it is necessary to export some of the information stored in the cache buffer. That is, the LRU algorithm is based on temporal locality.

However, when determining the priority of the internal memory block exchange of the cache buffer based solely on the chronological LRU algorithm, it replaces the place of the data that is unlikely to be cache hit or the data that is likely to be cache hit. The problem of occupying an internal memory block of a buffer may occur.

SUMMARY OF THE INVENTION The present invention provides a cache buffer management method for determining an internal memory block exchange priority of a cache buffer based on a folder hierarchy level of a file system and a disk drive using the same in order to solve the above problems. .

In order to achieve the above technical problem, a cache buffer management method according to the present invention includes determining whether a condition for replacing information stored in a cache buffer with new information occurs; In the case where a condition to replace the information stored in the cache buffer occurs, the priority is determined to replace the file data in preference to the folder data, and the hierarchical distance is calculated by calculating the hierarchical distance from the highest level folder among the folder data. Determining priority to preferentially replace remote subfolder data; And replacing the information stored in the cache buffer with new information according to the determined priority.

In the present invention, it is preferable to determine the priority by applying the LRU (Least Recently Used) algorithm to the folder data or the file data of the same level.

In accordance with another aspect of the present invention, a disk drive includes a disk for storing data; A cache buffer that temporarily stores data read from the disk or temporarily stores data to be written to the disk; And in the case where a condition to replace the data stored in the cache buffer occurs, the priority is determined to replace the file data in preference to the folder data, and the folder data is located at a lower level folder far from the top level folder. And a controller for determining priority to replace data first, and controlling to replace data stored in the cache buffer according to the determined priority.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The hard disk drive, which is a data storage device to which the cache buffer management method according to the present invention is applied, is composed of a combination of an electrical circuit and a head disk assembly (HDA) composed of mechanical components.

1 shows a configuration of a head disk assembly (HDA) 10 of a hard disk drive to which the present invention is applied.

The head disk assembly 10 includes at least one magnetic disk 12 that is rotated by the spindle motor 14. The disk drive 10 also includes a transducer 16 located adjacent to the surface of the disk 12.

The transducer 16 can read or write information on the rotating disk 12 by sensing and magnetizing the magnetic field of each disk 12. Typically transducer 16 is coupled to the surface of each disk 12. Although illustrated and described as a single transducer 16, it should be understood that this consists of a write transducer for magnetizing the disc 12 and a separate read transducer for sensing the magnetic field of the disc 12. Read transducers are constructed from Magneto-Resistive (MR) devices. The transducer 16 is also commonly referred to as a head.

The transducer 16 can be integrated into the slider 20. The slider 20 is structured to create an air bearing between the transducer 16 and the surface of the disk 12. The slider 20 is coupled to the head gimbal assembly 22. The head gimbal assembly 22 is attached to an actuator arm 24 having a voice coil 26. The voice coil 26 is located adjacent to the magnetic assembly 28 that specifies the voice coil motor 30 (VCM). The current supplied to the voice coil 26 generates a torque for rotating the actuator arm 24 relative to the bearing assembly 32. Rotation of the actuator arm 24 will move the transducer 16 across the disk 12 surface.

The information is typically stored in an annular track of the disc 12. Each track 34 generally includes a plurality of sectors. Each sector includes a data field and an identification field. The identification field is composed of a gray code identifying a sector and a track (cylinder). The transducer 16 is moved across the surface of the disc 12 to read or write information on other tracks.

The disk 12 is largely divided into a maintenance cylinder (generally referred to as MC) area and a user data area for storing information related to the disk drive. The maintenance cylinder area stores various information related to the disc drive, including defect list information and read / write optimization parameter information, and is an area that a user cannot access arbitrarily. The user data area is an area in which a user can substantially write or read data. The user data area of the disk 12 is divided into a plurality of zones, each of which is assigned a spare cylinder for replacing a defect sector.

2 shows an electrical system capable of controlling a hard disk drive. The electrical system includes a preamplifier 210, a write / read channel 220, a cache buffer 230, a controller 240, a read only memory (250A), a random access memory (RAM) 250B, a host interface 260. ) And voice coil motor (VCM) driver 270.

The controller 240 may be a digital signal processor (DSP), a microprocessor, a microcontroller, or the like. The controller 240 records for reading information from or writing information to the disk 12 according to a command received from a host device (not shown in the drawing) via the host interface 260. Control the read channel 220.

The controller 240 is also coupled to the VCM driver 270 which supplies a drive current to the voice coil 26. The controller 240 supplies a control signal to the VCM driver 270 to control the excitation of the voice coil motor and the movement of the transducer 16.

The ROM 250A stores firmware and various control data for controlling the disk drive. Of course, programs for executing the cache buffer management method according to the present invention shown in FIG. 3 are also stored.

When the power supply to the disk drive is supplied to the RAM 250B, various kinds of information related to the disk drive read from the maintenance cylinder position of the disk 12 are loaded.

First, the operation of a general disk drive will be described.

In the data read mode, the disc drive moves the transducer 16 to the target position of the disc 12 and then outputs the electrical signal detected by the transducer 16 from the disc 12 to the preamplifier 210. Amplify to facilitate signal processing. Then, the read / write channel circuit 220 encodes a signal read from the disk 12 into a digital signal in accordance with a sector pulse generated by the controller 240, converts the signal into a stream data, and temporarily stores it in the cache buffer 230. After storing the data, the data is transmitted to the host device (not shown) through the host interface 260.

Next, in the write mode, the disk drive receives data to be written from the host device (not shown) through the host interface 260 and temporarily stores the data in the cache buffer 230, and then cache cache 230. After converting the data read in from the write / read channel 220 to a binary data stream suitable for the write channel of the disc 12, the converter converts the read data into a write current amplified by the preamplifier 210 at the time a sector pulse is generated. Through the 16, the disc 12 is recorded.

The cache buffer 230 is composed of a plurality of internal memory blocks, and the data to be read and stored in the disk 12 corresponding to the main memory means in the READ mode or to be written to the disk 12 in the WRITE mode. Save it.

Personal computers using a disk drive as a storage device have a folder-type file system in the form of a tree. That is, as shown in FIG. 4, there are sub folders SUB Folders, which are lower level folders connected to the root folder, under a root folder, which is a top level folder, and another sub folder under each sub folder. Folders exist, and files are linked and stored in subfolders.

In the above, the folder also has data like a normal file, the contents of the folder data includes the creation time, size, and actual data location information of the files or folders included in the folder.

In order to access the file data in the file system as shown in FIG. 4, it is necessary to access the information from the top folder to the bottom folder to find the data location of the final destination file. That is, to access the object file, the folders in the path must be accessed.

Accordingly, it can be seen that in such a file system, storing the folder data in the cache buffer 230 for a long time rather than the file data during the data access process stored in the disk drive may increase the cache hit rate. Also, it can be seen that the cache hit ratio can be increased by keeping the folder closest to the top folder in the cache buffer 230 longer than the folder which is not.

In the present invention, when a part of information stored in the cache buffer 230 needs to be replaced, the internal memory block of the cache buffer 230 is exchanged in consideration of such file system characteristics.

Based on this principle, the controller 240 determines the priority to replace the file data preferentially to the folder data when a condition that needs to replace the data stored in the cache buffer 230 occurs, and the highest level among the folder data. The priority is determined to preferentially replace the folder data of a lower level far from the folder of the level, and control to replace the data stored in the cache buffer 230 according to the priority determined in this way. In particular, the priority is determined by applying a Least Recently Used (LRU) algorithm to file data or to folder data of the same level.

For example, a condition in which data stored in the cache buffer 230 needs to be replaced is that the size of the information capacity read from the recording medium according to the READ command due to no cache hit is generated, and thus the empty storage space of the cache buffer 230. Includes cases greater than the size of. Also, the size of data to be recorded on the recording medium according to the write command is also larger than the size of the empty storage space of the cache buffer 230.

Then, the cache buffer management method according to the present invention executed by the controller 240 will be described in detail with reference to the flowchart of FIG. 3.

The controller 240 determines whether a read command is received from the host device through the host interface 260 (S310).

When the read command is received, the controller 240 reads folder / file record information on the folder and file to be read from the disk 12 before accessing the folder or file data to be read (S320). . Folder / file record information includes folder / file name, folder / file creation date, folder / file size, and the like. Folder / file record information on disk 12 is generally stored in a different area from file data information. to be.

Next, the folder or file data is read according to the read command, and before accessing the disk 12 to read the data of the corresponding file, it is determined whether the folder or file data to be read is stored in the cache buffer 230. (S330). That is, it is determined whether a cache hit occurs.

If a cache hit occurs as a result of the determination in step 330 (S330), the corresponding folder or file data is read directly from the cache buffer 230 and transmitted to the host device through the host interface 260 without accessing the disk 12. (S340).

If the cache hit does not occur as a result of the determination in step 330 (S330), the disk 12 is accessed to read data of the corresponding folder or file (S350).

It is determined whether the size of the folder or file data read from the disk 12 exceeds the size (remaining size) of the empty storage space of the cache buffer 230 (S360).

If the size of the folder or file data read from the disk 12 does not exceed the size (remaining size) of the empty storage space of the cache buffer 230 as a result of the determination in step 360 (S360), the cache buffer 230 The folder or file data read from the disk 12 is stored in the remaining memory space at step S370.

If the size of the folder or file data read from the disk 12 exceeds the size (remaining size) of the empty storage space of the cache buffer 230 as a result of the determination in step 360 (S360), the file data is converted into folder data. The priority of the replacement of the internal memory block of the cache buffer 230 is determined to be replaced first, and the layer data is calculated from the folder of the highest level among the folder data to replace the lower folder data having a far higher distance. The priority of replacing the internal memory block of the cache buffer 230 is determined (S380). The priority is determined by applying a Least Recently Used (LRU) algorithm between file data or folder data of the same level.

As an example, assume that the root folder, subfolders 1-2, file # 1, subfolders 2-1, and subfolders 2-2 are stored in the cache buffer 230 in the tree structure file system as shown in FIG. The replacement priority in the cache buffer 230 is determined in order of file # 1, (subfolder 2-1, subfolder 2-2), subfolder 1-2, and root folder. Therefore, file # 1 is replaced first.

In the above example, assuming that file # 1 is not stored in the cache buffer 230, (subfolder 2-1 and subfolder 2-2) correspond to the replacement priority when considering the distance from the root folder. Since the subfolders 2-1 and 2-2 have the same hierarchical distance from the root folder, the LRU (Least Recently Used) algorithm is applied to the two folder data and replaced with the most recently unused data. Give priority.

Internal memory blocks of the cache buffer 230 are exchanged based on the priority determined according to step 380 (S380) (S390). That is, the data stored in the internal memory block of the cache buffer 230 to be replaced with the highest priority is deleted based on the priority determined according to the step 380 (S380), and the data is deleted from the disk 12 in the internal memory block from which the data is deleted. Save the read folder or file data.

As described above, in the present invention, the priority of the memory block exchange of the cache buffer 230 may be determined in consideration of the characteristics of the file system applied to the disk drive, thereby increasing the cache hit rate.

The invention can be practiced as a method, apparatus, system, or the like. When implemented in software, the constituent means of the present invention are code segments that necessarily perform the necessary work. The program or code segments may be stored in a processor readable medium or transmitted by a computer data signal coupled with a carrier on a transmission medium or network. Processor readable media includes any medium that can store or transmit information. Examples of processor-readable media include electronic circuits, semiconductor memory devices, ROMs, flash memory, erasable ROM (EROM), floppy disks, optical disks, hard disks, optical fiber media, radio frequency (RF) networks, Etc. Computer data signals include any signal that can propagate over transmission media such as electronic network channels, optical fibers, air, electromagnetic fields, RF networks, and the like.

Specific embodiments shown and described in the accompanying drawings are only to be understood as an example of the present invention, not to limit the scope of the invention, but also within the scope of the technical spirit described in the present invention in the technical field to which the present invention belongs As various other changes may occur, it is obvious that the invention is not limited to the specific constructions and arrangements shown or described. That is, it is a matter of course that the present invention can be applied not only to various disk drives including hard disk drives, but also to various kinds of data storage devices.

As described above, according to the present invention, in consideration of the characteristics of the file system of the tree structure applied to the disk drive, the priority of the exchange of the internal memory block of the cache buffer is replaced by the file data rather than the folder data. By determining based on the hierarchical distance from the folder at the top level, the effect of increasing the cache hit rate and shortening the data access time is generated.

Claims (5)

Determining whether a condition for replacing information stored in the cache buffer with new information occurs; In the case where a condition to replace the information stored in the cache buffer occurs, the priority is determined to replace the file data in preference to the folder data, and the hierarchical distance is calculated by calculating the hierarchical distance from the highest level folder among the folder data. Determining priority to preferentially replace remote subfolder data; And And replacing information stored in the cache buffer with new information according to the determined priority. The cache buffer management method of claim 1, wherein the priority is determined by applying a least recently used algorithm to each other of the folder data or the file data of the same level. The method of claim 1, wherein the condition that the information stored in the cache buffer needs to be replaced with new information is that a cache hit does not occur so that the size of the information capacity read from the recording medium according to the read command is larger than that of the empty storage space of the cache buffer. Cache buffer management method comprising a large case. A disk for storing data; A cache buffer that temporarily stores data read from the disk or temporarily stores data to be written to the disk; And In a case where a condition that replaces the data stored in the cache buffer occurs, the priority is determined to replace the file data in preference to the folder data, and the folder data at a lower level far from the folder at the highest level among the folder data. And a controller for determining a priority to first replace the data, and controlling to replace the data stored in the cache buffer according to the determined priority. The disk drive of claim 4, wherein the priority is determined by applying a Least Recently Used (LRU) algorithm to file data or to folder data of the same level.

Images