KR101681402B1 - Method for arranging data storage block of flash memory based on database table name - Google Patents

Abstract

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a data storage block arrangement method of a flash memory according to a database table name.
The method of the present invention includes a group setting step of setting data having the same database table name (DB table name) to be included in the same data group, and a data storing step of storing data included in the same data group in the same flash memory block And a control unit.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a flash memory,

The present invention relates to a semiconductor memory device, and more particularly, to a flash memory based data storage method.

Flash memory is a nonvolatile memory and is classified into NAND flash, which is similar to the characteristics of storage media such as NOR flash and hard disk, which have many characteristics of conventional random memory.

The internal structure of the NAND flash memory is composed of a plurality of blocks as shown in FIG. 1, and each block is composed of a plurality of pages. For example, one block may consist of 64 or 128 pages.

The NAND flash memory performs read and write operations on a page basis, and erase operations are performed on a block basis. The operation speed of each operation is different from each other. For example, the read operation is about 25 us, the write operation is about 250 us, and the erase operation is about 2,000 us, so that the speed of each operation is asymmetrical, If the operation is repeated, the operation speed of the memory may be slowed as a whole.

However, in the case of a hard disk drive, when an update occurs, the NAND flash memory can be overwritten in the same physical area of the existing data (in-place update). On the other hand, This is possible.

That is, the NAND flash memory is characterized in that new information can be recorded at the position only when the corresponding block is deleted. Due to this feature, it is impossible to immediately overwrite the location where the data was previously recorded, and the location information must be updated after out-place update.

Therefore, in order to improve the operation speed in the NAND flash memory, a technique of reducing the number of block deletion has been recognized as important.

On the other hand, since the number of times of writing / erasing in the flash memory is limited to a predetermined number of times (for example, 100,000 times), if the same block is used more than the corresponding number, the block can no longer be normally used (new writing action). However, in the conventional flash memory control method, a large number of read and write operations have been repeated in the process of garbage collection in the background operation.

Therefore, in order to extend the lifetime of the flash memory, there is a need for a technique for reducing the generation of fragments from the beginning to reduce the read / write operation in the background operation.

The inventor of the present invention has endeavored for a long time to solve this problem, and finally completed the present invention.

SUMMARY OF THE INVENTION It is an object of the present invention to provide an apparatus and a method for storing data of a flash memory that reduces data to be deleted at a similar time in consideration of a database table name property in the same block, thereby reducing additional background write operations required to move FRAGMENTS in a block Storage block allocation method.

It is still another object of the present invention to provide a method and apparatus for transferring meta data capable of estimating data deletion time through a host-device interface (hereinafter referred to as NVMe) such as NVM-Express To minimize fragmentation and minimize copying and deletion of data when garbage collection is performed.

On the other hand, other unspecified purposes of the present invention will be further considered within the scope of the following detailed description and easily deduced from the effects thereof.

According to another aspect of the present invention, there is provided a method of arranging data storage blocks in a flash memory, the method comprising: setting data whose deletion time is within a first deletion time range to a first data group; To a second data group; And

And a data storing step of storing data included in the first data group in the first flash memory block and storing data included in the second data group in the second flash memory block,

The deletion time is a time at which data stored in the flash memory is deleted from the memory block,

And the first deletion time range is present in time later than the second deletion time range.

According to another aspect of the present invention, there is provided a method of controlling a memory device, comprising: setting a storage period, which is a period from a time at which data is stored in a memory block to a time at which the data is stored, And a deletion time calculating step of calculating a deletion time by adding the storage period.

In a preferred embodiment, the group setting step of the present invention is characterized in that data having the same file extension is included in the same data group.

In a preferred embodiment of the present invention, before the group setting step, the data having the same file extension is set to have the same storage period, the deletion time calculation for calculating the deletion time by adding the storage period to the time at which the data is stored Further comprising the steps of:

In a preferred embodiment, the group setting step of the present invention sets data having an executable file extension (.exe) as a first data group and data having a temporary file extension (.tmp) as a second data group .

In a preferred embodiment, the present invention is characterized by further comprising a feedback step of storing a storage period log of data having the same file extension for a preset period to update a storage period or a deletion time.

In a preferred embodiment, the group setting step of the present invention sets data groups having different file extensions to the same data group when the storage period is within a preset range.

According to a preferred embodiment of the present invention, the group setting step of the present invention is characterized in that data having the same file extension and whose sequence number of the file extension is within a preset range is set to be included in the same data group .

In a preferred embodiment, the present invention further comprises a feedback step of storing a log of the storage period and the sequence number of the data having the same file extension for a preset period to update the storage period or the deletion time do.

In a preferred embodiment of the present invention, in the group setting step of the present invention, even if a file extension is the same for an extension specified in advance that there is no association between a file extension and a storage period, the sequence number of the file extension is stored in advance And is set to be included in the same data group only when it is within the set range.

In a preferred embodiment, the flash memory control method of the present invention is characterized by controlling a flash memory device executed according to an FTL algorithm through the NVMe protocol.

In a preferred embodiment of the present invention, the deletion time is included in metadata transmitted to the FTL according to the FTL algorithm through the NVMe protocol.

Conventionally, since a block is allocated according to a time when data is written and the block is arranged irrespective of the characteristics of data, in the course of garbage collection of a fragment generated as a result of use time as a background operation, However, in the present invention, since metadata that can estimate the data deletion time is transmitted and data having the same deletion time is stored in the actual memory in the same block, the fragment size is minimized And the data copying and deletion can be minimized when the garbage collection is performed.

In addition, according to the present invention as described above, the lifetime of a flash memory aging for a predetermined time or more can be extended (durability is improved).

In addition, according to the present invention, there is an effect that the writing speed can be improved in a flash memory used for a predetermined time or more.

In addition, data reading / writing and background GC competition are minimized to improve performance.

On the other hand, even if the effects are not explicitly mentioned here, the effect described in the following specification, which is expected by the technical features of the present invention, and its potential effects are treated as described in the specification of the present invention.

1 is a view for explaining a read / write operation of a flash memory according to the prior art.
2 is a diagram schematically illustrating an embodiment of a computer system using a data storage block arrangement method of a flash memory of the present invention.
3 is a diagram illustrating an embodiment of a software layer of a computer system that utilizes a method for arranging data storage blocks of a flash memory of the present invention.
4 is a diagram illustrating one embodiment of a software layer of a computer system that utilizes a data storage block placement method of a flash memory of the present invention.
5 is a diagram illustrating an alternative embodiment of FIG. 4 as a software layer of a computer system that utilizes a data storage block placement method of a flash memory of the present invention.
Figure 6 is an illustration of an alternate embodiment of Figure 5 as a software layer of a computer system utilizing a method of allocating data storage blocks of a flash memory of the present invention.
7 is a flowchart showing an embodiment of a data storage block arrangement method of the flash memory of the present invention.
It is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may obscure the subject matter of the present invention.

A flash memory device is a kind of nonvolatile memory. Flash memory devices provide faster data access performance than conventional hard disks and have very low volume and power consumption. In addition, flash memory devices are strong against external shocks and are widely used as storage devices for portable devices.

Flash memory has a certain level of lifetime, which can be determined based on the number of write / erase operations. On the other hand, the flash memory can not perform an overwrite operation due to its physical characteristics. That is, when a write operation is performed to a memory block in which data is written, the flash memory performs an erase operation before a write operation of the memory block.

System software called a Flash Translation Layer (FTL) is used between the file system of the host system and the NAND flash memory according to the characteristics of the NAND flash memory alone. The FTL hides the unique characteristics of the NAND flash memory to the host system's file system. It also uses the physical address of the NAND flash memory to store the logical address from the file system so that it can perform the same input / output operations as the hard disk. Mapping.

That is, the Flash Translation Layer (FTL) is software for efficiently managing the flash memory. The flash translation layer is included in the memory controller. The flash translation layer receives a logical address from the file system and converts it into a physical address of the flash memory device.

2 is a diagram schematically illustrating an embodiment of a computing system 1000 that utilizes a method for arranging data storage blocks of a flash memory of the present invention.

As can be seen in Figure 2, the computing system 1000 includes a host 1100, a memory controller 1200, and a flash memory 1300. Each of host 1100, memory controller 1200, and flash memory 1300 of computing system 1000 may be comprised of separate chips, modules, or devices, or may be included within a single device. Also, the memory controller 1200 and the flash memory 1300 may be implemented as a single device, and may be used in connection with the host 1100.

Host 1100 may send a read or write request to memory controller 1200 using an application or file system. The memory controller 1200 may control the operation of the flash memory 1300 in response to a read or write request from the host 1100. [ For example, in response to a read or write request from the host 1100, the memory controller 1200 may control the write, read, and erase operations of the flash memory 1300.

In the preferred embodiment, the host 1100 and the memory controller 1200 can exchange data based on a predetermined protocol. For example, the host 1100 and the memory controller 1200 may be connected to each other via a universal serial bus (USB) protocol, a multimedia card (MMC) protocol, a peripheral component interconnection (PCI) protocol, a PCI- At least one of various interface protocols such as Advanced Technology Attachment protocol, Serial-ATA protocol, Parallel-ATA protocol, small computer small interface (SCSI) protocol, enhanced small disk interface (ESDI) protocol, and Integrated Drive Electronics You can exchange data based on one.

Memory controller 1200 may include a processor 1210 and a RAM 1220. The processor 1210 can control all operations of the memory controller 1200. For example, processor 1210 may drive a Flash Translation Layer (FTL; shown in FIG. 3).

The RAM 1220 may be an operational memory of the memory controller 1200. The RAM 1220 may be a buffer memory between the host 1100 and the flash memory 1300. The RAM 1220 may include a mapping table 1221 (Mapping Table). The mapping table 1221 is information for a conversion operation of the flash conversion layer (FTL). The mapping table 1221 includes information on a mapping relationship between a logical address of a file organized by a file system (shown in FIG. 3) and a physical address of the corresponding flash memory 1300 do. In a preferred embodiment, when a free memory block is allocated as a write memory block, the memory controller 1200 may update the mapping table 1221. [

The flash memory 1300 may perform writing, reading, or erasing operations of the memory blocks under the control of the memory controller 1200. The flash memory 1300 may include a plurality of memory blocks including a plurality of pages. The flash memory 1300 may include a user area 1310 and a meta area 1320. The user area 1310 is an area for storing user data. The user area 1310 may comprise a plurality of memory blocks included in the flash memory 1300. Each of the plurality of memory blocks may be composed of a plurality of pages. Each of the plurality of pages includes a data area for storing user data and a spare area for storing additional information. The additional information may include an error correcting code (ECC), a logical address of data stored in the data area, bad block information, and the like. Illustratively, some of the spare areas may include the physical address of the next write memory block to be allocated.

The meta area 1320 includes a main mapping table 1321. The main mapping table 1321 includes information on a mapping relationship between a logical address received from the host 1100 and a corresponding physical address of the flash memory 1300.

The memory controller 1200 can read the main mapping table 1321 and store the read main mapping table 1321 in the mapping table 1221 of the RAM 1220. [ Illustratively, the memory controller 1200 may allocate a new write memory block in response to a write request received from the host 1100. The memory controller 1200 can update the mapping table 1221 based on the mapping relationship between the physical address of the memory block allocated as the new write memory block and the corresponding logical address.

Illustratively, the meta area 1320 may be comprised of one or more memory blocks of a plurality of memory blocks included in the flash memory.

3 is a diagram of an embodiment of a software layer of a computer system that utilizes a method for arranging data storage blocks of a flash memory of the present invention.

2 and 3, the software layer of the computing system 1000 may include an application 1110, a file system 1120, a flash translation layer 1230, and a flash memory 1300.

Application 1110 refers to various application programs running in the computing system 1000. For example, application programs 1120 may include a text editor, a web browser, a video player, a game program, and the like.

The file system 1120 organizes the file or data when the file or data is stored in the flash memory 1300. For example, the file system 1120 may provide a logical address to the memory controller 1200 in response to a write request. The file system 1120 may have a different form according to an operating system (OS) of the host 1100. Illustratively, the file system 420 may include a file allocation table (FAT), FAT32, NTFS, HFS, JSF2, XFS, On-Disk Structure- 5, UDF, ZFS, Unix File System, ext2, ext3, ext4, ReiserFS, Reiser4, ISO 9660, Gnome VFS, BFS, or WinFS. By way of example, the file system 1120 may be powered by the host 1100.

The flash translation layer 1230 can provide an interface between the host 1100 and the flash memory 1300 such that the flash memory 1300 can be used efficiently. The flash conversion layer 1230 receives a logical address generated by the file system 1120 and converts the logical address into a physical address usable in the flash memory 1300. The flash conversion layer 1230 manages such address conversion through a mapping table 1221. [ Illustratively, the flash translation layer 1230 may be driven by the memory controller 1200.

4 is a diagram illustrating one embodiment of a software layer of a computer system that utilizes a data storage block placement method of a flash memory of the present invention.

4 illustrates a method of arranging a data magnetic field block of a flash memory in an embodiment using a NVMe file system in a computer system. Referring to FIG. 4, the NVMe file system provides an NVMe interface to provide an additional interface. Meta data management logs data life and creates metadata. The NVMe block device layer is made using open source and provides an NVMe interface.

The FTL may include a Sector Translation Layer (STL), a Block Management Layer (BML), and a Low Level Driver (LLD). The STL performs mapping, garbage collection, and wear leveling using metadata. BML performs bad block management and error handling. LLD provides a flash interface.

One embodiment of the data storage block arrangement method of the flash memory of the present invention may include the deletion time of the data to be stored in the metadata transmitted to the FTL according to the FTL algorithm through the NVMe protocol. To this end, the file system and FTL can be changed / improved to create and transfer the changed metadata.

The erase time means the time at which data stored in the flash memory is deleted from the memory block. The data storage block allocation method of the present invention prevents unnecessary data erasure operations from being repeated by storing the same data in the same block within a predetermined range of erasure time.

The deletion time passed from the NVMe file system to the FTL may include all of the ways in which the application's I / O is delivered without change or is delivered via I / O changes.

First, with reference to FIG. 4 and FIG. 5, an embodiment for transferring the deletion time without changing the I / O of the application will be described.

First, the application I / O stores the file in an unchanged application. For example, an event occurs that saves a file called aaa.doc.

In the NVMe file system, information including the deletion time is included in the metadata through the conventional I / O and transmitted to the metadata manager according to the NVMe protocol. For example, information about the extension "doc" serving as a criterion for determining the deletion time is added to the metadata of the NVMe protocol and transmitted.

The metadata manager transforms the metadata forwarded from the file system so that it can be processed by the FTL. For example, metadata is converted by designating a deletion time category number (category number) corresponding to the extension "doc" according to a predetermined deletion time analysis result or an update result through feedback based on log analysis. The category number can be specified, for example, as follows.

0: Ignore metadata

1: very short

2: Deletion time arrives within 1 day

3: Deletion time arrives within 1 week

4: Deletion time arrives within 1 month

5: Very Kim

The processed information is transferred to the FTL through block I / O according to the NVMe protocol. The STL analyzes the transferred metadata and maps the data having the same deletion time to be stored in the same block.

Next, with reference to FIG. 4 and FIG. 6, an embodiment for changing the I / O of the application and transmitting the deletion time through it will be described. The application passes the metadata with the deletion time through the extended I / O interface. The NVMe file system delivers the delivered metadata according to the NVMe protocol. The metadata manager transforms the metadata delivered by the application through the extended interface into a promised format for processing by the FTL. The STL maps the data having the same deletion time to be stored in the same block using the received metadata.

7 is a flowchart showing an embodiment of a data storage block arrangement method of a flash memory of the present invention.

As shown in FIG. 7, the data storage block allocation method of the present invention may perform the step of calculating the deletion time in advance (2100). The calculated erasure time in the preferred embodiment may be stored in software contained within the FTL.

In one embodiment, the deletion time calculation step 2100 sets the storage period, which is a period from the time when data is stored in the memory block to the time when it is deleted, according to the characteristics of the data, adds the storage period to the time at which the data is stored The deletion time can be calculated. The inventors of the present invention have found that data storage periods differ depending on the characteristics of data. For example, data such as program executable files, images, and movies are often executed but rarely deleted. On the other hand, temporary files have a short storage period. Thus, the present invention classifies the storage periods according to the characteristics of the data.

In another embodiment, the deletion time calculation step 2100 may set the data having the same file extension to have the same storage period, and calculate the deletion time by adding the storage period to the time when the data is stored. In the above embodiment, the storage period is classified according to the characteristics of the data. In the modified example, the storage period is calculated using the file extension.

In a preferred embodiment, the present invention can estimate the storage period of data characteristics or extensions according to the following table.

Classification Data Characteristics Extension Example Temporary files Short storage period .tmp, swp, etc. Application data file The data file used for each program. Different storage period by program characteristics .doc, .hwp, .pptx, etc. Archive file Data archive file. Very long storage period .zip, .tar, etc. Multi-media file Image, video, etc. The storage period is quite long and can not be modified. .jpg, .avi, .gif, etc. Executable Executable files. Modification is not good and storage period is long .exe, .com, etc.

Next, a group setting step 2200 for setting the data whose deletion time is within the first deletion time range to the first data group and setting the data whose deletion time is within the second deletion time range to the second data group have. The deletion time is a time at which the data stored in the flash memory is deleted from the memory block, and the first deletion time range exists in a time later than the second deletion time range. That is, data to be deleted at a first deletion time of a similar time period is set as a first data group, and data to be deleted at a second deletion time that precedes the first data group is separated into a second data group.

The present invention includes various embodiments as to what criteria are grouped into the same group in the group setting step 2200. [

In one embodiment, the group setting step 2200 may set data having the same file extension to be included in the same data group. For example, data having an executable file extension (.exe) can be set as a first data group, and data having a temporary file extension (.tmp) can be set as a second data group. If the file has the same extension, the expected deletion time may be the same. In this embodiment, since the group can be set only by the extension without actually calculating the storage period and the deletion time, the step 2100 can be omitted, which is convenient.

In another embodiment, the group setting step 2200 may set the data group having the different file extensions to the same data group if the storage period is within the preset range. In such an embodiment, it may be more useful to go through step 2100 in order to select extensions similar in actual storage duration.

In another embodiment, data having the same file extension and having a sequence number of the file extension within a preset range can be set to be included in the same data group. In this embodiment, the concept of a sequence number is introduced. The sequence number is a number for identifying the storage time. If the same data attribute or the same extension is used for the same file, the sequence number can be included in different data groups. Thus, the sequence number may increase with the number of files or over time. However, it is more efficient if the number is incremented in time units than when it is increased according to the number of files. How to determine the unit time to increase the sequence number can be configured differently depending on the characteristics of the hardware.

In another embodiment, even if the file extension is the same for an extension specified in advance that there is no association between the file extension and the storage period, only when the sequence number of the file extension is within the preset range, As shown in FIG. For example, when editing a moving image, a command to correct a large-capacity file at a specific portion is repeated. In the case of video files, the file size is very large, so one file is stored in many blocks. In a typical case, such a moving picture file is stored and deleted at once. However, in the moving picture editing process, a part of the file is deleted or changed. Therefore, in such a case, it may be meaningless to arrange blocks in units of extensions. Therefore, in this case, even if they have the same extension, block arrangement is performed based on the same sequence number.

Next, a data storage step of storing data included in the first data group in the first flash memory block and storing data included in the second data group in the second flash memory block is executed (step 2300). The present invention has the effect of preventing unnecessary data deletion operations from being repeated because data having the same deletion time is stored in the same flash memory block.

Next, the data storage block allocation method according to the present invention may further include a feedback step 2400 of storing a storage period log of data having the same file extension for a preset period to update a storage period or a deletion time . The present invention can be more finely optimized through feedback on erasure time. In another embodiment, the present invention may include a feedback step of updating a storage period or a deletion time by storing a log of a storage period and a sequence number of data having the same file extension for a predetermined period.

The present invention can also be applied to a data storage block arrangement method of a flash memory for storing a database file. In this case, the present invention may include a group setting step of setting data having the same database table name (DB table name) to be included in the same data group; And a data storing step of storing the data included in the same data group in the same flash memory block. Files with the same database table name have similar delete times.

Also, in the preferred embodiment, the group setting step of the above embodiment may set the data having the same database table name and having a sequence number within a preset range to be included in the same data group.

The scope of protection of the present invention is not limited to the description and the expression of the embodiments explicitly described in the foregoing. It is again to be understood that the present invention is not limited by the modifications or substitutions that are obvious to those skilled in the art.

Claims (10)

delete The data having the same database table name is set to have the same storage period (the period from the time the data is stored in the memory block to the time the data is deleted), and the storage period is added to the time at which the data is stored, A deletion time calculation step of calculating a deletion time;
Setting the data whose deletion time is within the first deletion time range to the first data group and setting the data whose deletion time is within the second deletion time range to the second data group; And
Storing data included in the first data group in a first flash memory block and storing data included in a second data group in a second flash memory block that is a block different from the first flash memory block ,
Wherein the first deletion time range is present in a future time than the second deletion time range in the future. 3. The method of claim 2,
Wherein data having the same database table name and having a sequence number within a range set in advance is included in the same data group. The method of claim 2,
Further comprising a feedback step of storing a storage period log of data having the same database table name and updating the storage period or the deletion time during a preset period of time. Block placement method. The method according to claim 2,
Wherein the same data group is set as a data group even if the data having different database table names are within a preset range. The method according to claim 2,
And setting the data having the same database table name and a sequence number of the database table name within a preset range to be included in the same data group. How to place data storage blocks. delete The method of claim 2,
Wherein the data storage block arrangement method of the flash memory controls the flash memory device executed according to the FTL algorithm through the NVMe protocol. The method of claim 2,
Wherein the data on the deletion time is included in the metadata transmitted to the FTL according to the FTL algorithm through the NVMe protocol. The method of claim 2,
The data on the deletion time is included in the metadata transmitted to the FTL according to the FTL algorithm through the NVMe protocol, and the metadata is transferred through the added I / O interface of the application. A method for arranging data storage blocks in a flash memory.

Images