KR20090042077A - File Management System and Method - Google Patents

Abstract

A nonvolatile memory based file management system and method are provided. A storage device of the present invention of the present invention is a block device for storing data, and non-volatile random access memory (byte-addressable), and non-volatile random access memory for storing metadata for controlling the block device (non- volatile random access memory (NVRAM), which can reduce the time required to mount.

Description

File Management System and Methods {FILE MANAGEMENT SYSTEM AND METHOD}

The present invention relates to a file management system and a file system, and more particularly, to a file management system and a file system based on a non-volatile memory.

The data storage device may include a magnetic disk, a semiconductor memory, and the like, and since the data storage device has different physical characteristics for each type, a management system corresponding to the physical characteristics is required.

As a conventional data storage, magnetic disks have been widely used. Magnetic disks have an average read / write time of a few milliseconds per kilobyte. In addition, the magnetic disk has a characteristic that the read / write time is different because the time that the arm arrives varies depending on the physical location where the data is stored.

Recently, nonvolatile memory devices, which have shorter read / write times, consume less power, and occupy smaller volumes than magnetic disks, are rapidly replacing magnetic disks. This is possible due to the large capacity of the nonvolatile memory device.

The nonvolatile memory device is a semiconductor memory device capable of electrically reading, writing, and erasing and maintaining programmed data even in a state in which there is no power supply. In addition to writing, the process of storing data for nonvolatile memory devices is also called programming.

Non-volatile memory devices are widely used in digital cameras, cellular phones, personal data assistants (PDAs), laptop computers, and the like, and store large amounts of data in small areas on semiconductor chips. Can be stored.

Among non-volatile memory devices, NAND flash memory is widely used as a mass storage device. NAND flash memory features tens of microseconds per kilobyte of read time and hundreds of microseconds of write time. In addition, NAND flash memory generally has an erase time different from the write time. Since the erase time is several milliseconds, a method of simultaneously erasing a plurality of memory cells is widely used to reduce the erase time for the entire data. . In this case, a unit of a plurality of memory cells that are simultaneously erased is called an erase unit.

As described above, the nonvolatile memory including NAND flash memory has asymmetrical read / write / erase characteristics in which the time required for read / write / erase is all different from the conventional magnetic disk. In addition, since the nonvolatile memory electrically accesses the data, the location where the data is stored has little effect on the read / write / erase characteristics.

In the conventional magnetic disk, the process of erasing data is included in the data writing process, and thus, a separate time is not required. Therefore, if data needs to be updated, the updated data may be overwritten to a position of the previous data.

However, in the nonvolatile memory, the data erasing process and the data writing process are separated, so that a separate long time is required for erasing the data. Therefore, if the conventional magnetic disk method is applied to the nonvolatile memory, the data is updated. This takes a long time. In particular, it is known that the update time becomes longer when the data is randomly accessed and the data is randomly updated.

In addition, a file system is widely used to manage such mass storage-based data. However, in order to use a file system, first, a process of reading management information called metadata stored separately in storage for each file system and decrypting the information so that the system knows it is necessary. However, most of these processes require a very long time due to the large capacity of the storage device. In the case of a hard disk, the initial driving time is long, and in the case of a flash memory, a wear leveling process is performed to extend the service life in consideration of the read / write characteristics of the flash memory.

In the present specification, a more efficient file management system is proposed for a nonvolatile memory based file system.

SUMMARY OF THE INVENTION The present invention has been made to solve the problems of the prior art as described above, and reduces the time required for mounting by directly mapping metadata to non-volatile random access memory (NVRAM). For the purpose of

The present invention also aims to shorten the time for accessing and updating the metadata by storing the metadata in NVRAM as an in-memory data structure.

Another object of the present invention is to provide a method for managing a storage device without changing a conventional operating system (OS) by using byte-addresable NVRAM as a system memory.

In order to achieve the above object and to solve the problems of the prior art, the storage device of the present invention is a block device for storing data, and addressable in units of bytes, the ratio of storing the metadata for controlling the block device And a volatile random access memory.

In addition, the file management system according to another aspect of the present invention is a block device for storing data, addressable in units of bytes, a nonvolatile random access memory for storing metadata for controlling the block device, and the non- And a processor directly connected to the volatile random access memory and controlling the block device using the metadata.

In addition, the file management method according to another aspect of the present invention comprises the steps of storing the meta data for controlling the block device in a non-volatile random access memory as an in-memory data structure, and the file system information and the metadata of the block device Controlling the block device by mapping.

The present invention has been made to solve the problems of the prior art as described above, and reduces the time required to mount by directly mapping the metadata to non-volatile random access memory (NVRAM) .

In addition, the present invention shortens the time to access and update metadata by storing the metadata in NVRAM as an in-memory data structure.

In addition, the present invention provides a method of managing a storage device without changing a conventional operating system (OS) by using byte-addresable NVRAM as a system memory.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited or limited by the embodiments. Like reference numerals in the drawings denote like elements.

In general, storage for storing data is also called a block device. A block device refers to a non-volatile storage device such that a hard disk drive (HDD) or a flash memory does not lose stored data even when power is cut off. In contrast, a memory having a smaller capacity and a faster operating speed than a block device is sometimes referred to as system memory.

The system memory is generally a volatile memory in which all stored data is lost when the power is cut off, such as a dynamic random access memory (DRAM). In addition, the system memory generally has an addressable structure in units of bytes or words, and the structure may operate faster than the block device.

In general, in order to use a block device in an operating system (OS), a file system must be built in the block device. Block devices have pure data storage for the file system and metadata storage for indexing data. To access the data on the block device, the file system loads metadata on the block device onto system memory.

Flash memory, widely used as a block device, stores and manages data in blocks that are simultaneously erased. On the other hand, since metadata needs to be updated in byte units or word units as required by an operating system (OS), it is necessary to load metadata onto system memory. Thus, metadata is loaded onto system memory. Sometimes called the process of mounting.

Hard disks, which are widely used as block devices, also store and manage data in sectors or clusters, and thus require metadata mounting.

In this case, the metadata mounted on the system memory is not copied as it is stored in the block device, but is changed into an in-memory data structure suitable for an operating system and stored.

Examples of a file system used when the block device is flash memory include YETFS, YAFFS1, YAFFS2, Journaling Flash File System (JFFS), or JFFS2.

In the conventional file system using a flash memory (block device), metadata is stored only in the flash memory (block device). In general, as a method for extending the service life of a flash memory, data stored by a process called wear-leveling is distributed. If the metadata stored in the flash memory has been distributed by wear leveling, the operating system needs to scan the entire flash memory area to retrieve the stored metadata.

The retrieved metadata is changed into an in-memory data structure to be mounted in system memory.

Recently, due to the rapid development of flash memory manufacturing technology, a large capacity flash memory device has been developed. As the amount of flash memory increases, the time taken to retrieve stored metadata increases. Since the operating system requires the process of retrieving the metadata and mounting it on the system memory in order to manage the block device (flash memory), the increase in the retrieval time of the metadata significantly degrades the performance of the entire file system.

On the other hand, non-volatile random access memory (NVRAM), which is non-volatile and byte-addressable, has been developed.

Examples of NVRAM include FeRAM (Ferroelectric RAM), Magnetic RAM (MRAM), and Phase Change RAM (PRAM). These features are byte- or word-addressable, Even if blocked, the stored data is not lost.

NVRAM's read / write speed is still slower than conventional DRAM, but it's very fast compared to flash memory. On the contrary, NVRAM's capacity is still very small compared to flash memory.

The present specification proposes a method for replacing an existing system memory with NVRAM and a replaced storage device, which is possible due to the byte-addressable characteristics of NVRAM.

NVRAM is directly linked to the processor's memory address, which can be mapped to the processor's memory map, and the processor can read or write NVRAM in bytes as if it were accessing system memory. In addition, unlike DRAM, which is widely used as a system memory, NVRAM does not lose stored data even when the power is cut off, so that when the power is applied again, previously written data can be reused.

1 is a diagram illustrating a storage device 100 according to an embodiment of the present invention.

Referring to FIG. 1, the storage device 100 includes an NVRAM 110 and a block device 120.

NVRAM 110 is byte-addressable and stores meta data for controlling block device 120.

The block device 120 is a device that stores data, and in some embodiments, may be a flash memory or a hard disk. In this case, the flash memory may be addressable in units of blocks, and the hard disk may be addressable in units of sectors.

In some embodiments, the block device 120 may have only a pure data storage area for storing data. In this case, the NVRAM 110 may store the meta data as an in-memory data structure. In-memory data structures are data structures that have been modified to suit the operating system.

According to another exemplary embodiment, the block device 120 may include a pure data storage area and a metadata storage area for storing data. In this case, the second metadata in which the metadata is compressed may be stored in the metadata storage area.

According to the storage device 100, there is no need to scan the block device 120 to collect metadata and to mount the collected metadata on system memory as an in-memory data structure. Since the NVRAM 110 is nonvolatile, the stored metadata is not lost even when the power is cut off, and when the power is applied again, the stored metadata may be used. In addition, since the NVRAM 110 is byte-addressable, the NVRAM 110 may be updated in byte units even when updating metadata.

The mounting process of the storage device 100 may be completed by a simple process of mapping metadata of the NVRAM 110 to an in memory data structure of a file system that operates the block device 120.

2 is a diagram illustrating a file management system 200 according to an embodiment of the present invention.

The file management system 200 includes a processor 210, an NVRAM 220 and a block device 230.

The block device 230 stores data, and in some embodiments, may be a flash memory or a hard disk.

NVRAM 220 is byte-addressable and stores metadata for controlling block device 230.

The processor 210 is directly connected to the NVRAM 220 and controls the block device 230 using metadata. The processor 210 reads metadata stored in the NVRAM 220. The block device 230 is controlled by updating.

The processor 210 may control the block device 230 by mapping file system information and metadata of the block device 220.

Meta data is stored as an in-memory data structure that is directly available to the operating system on which file management system 200 operates.

The file management system 200 is an improvement over the prior art in which the block device 230 takes a long time to mount the metadata when the block device 230 has a large capacity.

The file management system 200 may shorten the time required for mounting the meta data to secure availability quickly.

The file management system 200 may shorten the time required for mounting because the type of metadata stored in the system memory is different from the type stored in the block device 230. In the file management system 200, the metadata is stored in the NVRAM 220 as an in-memory data structure, so that the meta data conversion process is unnecessary.

The file management system 200 may store metadata only in the NVRAM 220 to increase the efficiency of metadata management. The prior art has a problem of storing meta data redundantly in system memory and block device 230.

The file management system 200 stores the metadata in the byte-addressable NVRAM 220, so that the metadata can be quickly updated in byte units.

The file management system 200 may increase data management efficiency by separating a space for storing metadata and a pure data storage area for storing data. Meta data is stored in NVRAM 220 and data is stored in block device 230.

File management system 200 may replace conventional system memory with NVRAM 220 that stores metadata. The processor 210 is coupled to the byte-addressable NVRAM 220 via a system bus.

If the address space utilized by the file management system 200 is a 32-bit address space, a portion of the 32-bit address space may be allocated to the NVRAM 220 by the size of the NVRAM 220.

3 is a diagram illustrating an example of an operation of the file management system 300 according to an exemplary embodiment.

Referring to FIG. 3, the file management system 300 includes a data storage area 310, an in memory data structure metadata storage area 320, a file system driver 330, and an operating system 340.

The file management system 300 stores and manages the object structure and the Tnode structure, which are in-memory data structures of the object header and tag information, which are meta data of YAFFS, in FRAM (MRAM PRAM), which is byte-addressable NVRAM.

At this time, the NVRAM corresponds to the in-memory data structure metadata storage area 320.

When the operating system 340 accesses the data storage area 310, if the operating system 340 sends a mount command, the YAFFS file system driver 330 simply transfers the NVRAM Object Structure and Tnode Structure to the driver structure. Complete the mount operation by mapping.

All subsequent operations are performed in a similar manner as in conventional YAFFS.

4 is a flowchart illustrating a file management method according to an embodiment of the present invention.

Referring to FIG. 4, the file management method stores meta data for controlling a block device in NVRAM as an in memory data structure (S410).

The file management method maps file system information and metadata of the block device (S420).

The file management method controls the block device by mapping file system information of the block device and the meta data (S430).

According to an exemplary embodiment, the file management method may update metadata stored in NVRAM in units of bytes.

According to an embodiment, the file management method may copy metadata from the block device to NVRAM. At this time, the file management method may erase the meta data from the block device.

The file management method according to the present invention can be implemented in the form of program instructions that can be executed by various computer means and recorded in a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the media may be those specially designed and constructed for the purposes of the present invention, or they may be of the kind well-known and available to those having skill in the computer software arts. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, such as floppy disks. Magneto-optical media, and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

As described above, the present invention has been described by way of limited embodiments and drawings, but the present invention is not limited to the above embodiments, and those skilled in the art to which the present invention pertains various modifications and variations from such descriptions. This is possible.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined not only by the claims below but also by the equivalents of the claims.

1 is a diagram illustrating a storage device 100 according to an embodiment of the present invention.

2 is a diagram illustrating a file management system 200 according to an embodiment of the present invention.

3 is a diagram illustrating an example of an operation of the file management system 300 according to an exemplary embodiment.

4 is a flowchart illustrating a file management method according to an embodiment of the present invention.

<Explanation of symbols for the main parts of the drawings>

110: nonvolatile random access memory

120: block device

Claims (11)

A block device for storing data; And Non-volatile random access memory capable of addressing in units of bytes and storing metadata for controlling the block device Storage device comprising a. The method of claim 1, The block device has only a data storage area for storing the data, And the nonvolatile random access memory stores the metadata as an in-memory data structure. The method of claim 1, The block device comprises a metadata storage area and a data storage area, The metadata storage area, the storage device characterized in that the second metadata is compressed is stored. The method of claim 1, The block device A nonvolatile memory addressable in block units. A block device for storing data; A nonvolatile random access memory capable of addressing in units of bytes and storing meta data for controlling the block device; And A processor directly connected to the nonvolatile random access memory and controlling the block device using the metadata; File management system comprising a. The method of claim 5, And the processor controls the block device by mapping file system information and metadata of the block device. The method of claim 5, And the metadata is stored as an in-memory data structure that is directly available to an operating system on which the file management system operates. Storing meta data controlling the block device as a in-memory data structure in a nonvolatile random access memory; And Controlling the block device by mapping file system information and the metadata of the block device File management method comprising a. The method of claim 8, Updating the metadata stored in the nonvolatile random access memory in units of bytes. File management method comprising a further. The method of claim 8, Copying the metadata from the block device to the nonvolatile random access memory; And Erasing the meta data from the block device File management method that includes more. A computer-readable recording medium in which a program for executing the method of any one of claims 8 to 10 is recorded.

Images