Patent Application
20090157756
This application is related to commonly assigned U.S. Patent Application ______, (Attorney Docket Number HSJ9-2007-0208-US1), filed concurrently herewith, to Marco Sanvido, which is incorporated by reference herein.
The present invention relates to data storage systems, and more particularly, to a file system that stores files in multiple different data storage media.
A hard disk drive is a type of data storage device. A hard disk drive stores data onto the surface of one or more hard disk platters as a magnetic image. Other types of data storage devices include flash memory devices and optical disk drives, such as CD and DVD drives.
According to some embodiments of the present invention, a host system includes a file system and a processor for executing the file system. The file system stores a first portion of a file in a first data storage medium and a second portion of the file in a second data storage medium based on an intrinsic value of at least a part of the file. The first and the second data storage media are different types of data storage media.
According to other embodiments of the present invention, a file system stores a first file in a first data storage medium based on an intrinsic value of the first file. The file system stores a second file in a second data storage medium based on an intrinsic value of the second file. The file system dynamically moves the second file from the second data storage medium to the first data storage medium in response to a change in the intrinsic value of the second file. The first and the second data storage media are different types of data storage media.
Various objects, features, and advantages of the present invention will become apparent upon consideration of the following detailed description and the accompanying drawings.
A file system is a technique for storing and organizing computer files to facilitate the process of locating the files. File system 112 can be used to manage data blocks that are stored on data storage system 102. The file system 112 organizes the data blocks into files and directories. The file system 112 also keeps track of which data blocks belong to which file and which data blocks are not being used.
Data storage system 102 includes a controller 103, a first data storage medium 104, a second data storage medium 105, and a third data storage medium 106. Controller 103 is typically fabricated on an integrated circuit. Controller 103 processes read and write commands from host system 101. Controller 103 also communicates with each of the data storage media 104-106 through one or more communications channels. Controller 103 causes data to be read from and written to data storage media 104-106 in response to read and write commands from host system 101.
Each of the data storage media 104-106 is a non-volatile data storage medium. In one embodiment, data storage system 102 is a single data storage device that has 3 different types of data storage media, such as, for example, an optical disk, a magnetic disk, magnetic tape, and non-volatile semiconductor memory.
Data storage systems 102 and 202 are configured to map logical block addresses (LBAs) to physical addresses in different types of non-volatile data storage media 104-106. The physical addresses typically have different numerical values than the LBAs, but the physical addresses and the LBAs can have the same numerical values. In a hard disk drive, the physical addresses can be cylinder head sector numbers.
Data storage media 104-106 in
According to an embodiment, a file system automatically determines which data storage medium to use for storing each portion of a file based on the intrinsic value of the file or based on the intrinsic value of one or more portions of the file. A file system can, for example, select data storage media for storing portions of a file based on intrinsic values of a file such as an expected access time for portions of the file, an expected access rate for portions of the file, an expected sequential access time or rate for portions of the file, the value of the data or code in a portion of the file, the metadata sectors of the file, or a desired reliability of a portion of a file.
As a specific example, a file system and/or software application (e.g., a database application) may require a fast random access time and/or a fast random access rate to the data or code stored in a portion of a file. In this example, the file system stores the data or code in that portion of the file in Flash memory to provide a fast random access time and a fast random access rate to that data or code. One or more other portions of the file that do not require fast random access can be stored, for example, in a hard disk, in an optical disk, or in magnetic tape.
As another specific example, a file system or software application may require a fast sequential access time to the data or code in a portion of a file. Sequential access refers to reading or writing sectors in sequential order, i.e., a set of sectors that are adjacent to each other as written on the data storage medium. In this example, the file system stores the data or code in the portion of the file requiring fast sequential access in a hard disk or in an optical disk. One or more other portions of the file that do not require fast sequential access can be stored, for example, in Flash memory, or in magnetic tape.
As yet another specific example, a file system or software application may require a high degree of reliability for valuable data or code in one or more portions of a file. In this example, the file system stores the most valuable portions of the file in a highly reliable data storage medium. One or more other portions of the file can be stored in a less reliable storage medium.
As yet another specific example, the metadata portion of a file may have a different intrinsic value than the data portion of a file. The metadata portion of a file may be used for limited purposes, for example, to store the file's last access time. In this example, the file system stores the metadata portion 201A of the file 201 in a first data storage medium 104 and the data portions 201B and 201C of the file 201 in different data storage media. The file system stores the second portion 201B and the third portion 201C of the file in data storage media 105 and 106, respectively. The first data storage medium 104 can be a slower, less expensive medium, because the metadata portion 201A of the file is not accessed as often as the data portions 201B and 201C of the file.
According to another embodiment, a file system stores highly accessed parts of a file (e.g., the code portion of a Windows DLL executable file) in a faster, more expensive data storage medium. DLL stands for dynamic-link library. The file system stores the other parts of the file that are rarely accesses (e.g., the symbols in a Windows DLL executable file) in cheaper and slower media. These examples are provided for the purpose of illustration and are not intended to limit the scope of the present invention.
In some embodiments, the file system assigns different portions of a file to ranges of logical block addresses (LBAs) that map to different types of data storage media. The LBAs map to physical addresses in each of the data storage media. The physical addresses correspond to units of storage space in a particular data storage medium. Each data storage medium 104-106 has a physical address assigned to each of unit of storage space. Each physical address assigned to a unit of storage space in a data storage medium is unique with respect to the other physical addresses assigned to other units of storage space in that data storage medium.
Host 101 sends a range of LBAs to the data storage system with each read command and each write command to access data from data storage media 104, 105, and 106. After the data storage system receives a command and LBAs from host 101, a controller accesses the physical addresses in one or more of the data storage media 104-106 corresponding to the LBAs received from the host 101. The controller then reads data from or writes data to the one or more mapped data storage media 104-106. Further details of a data storage system that stores data in multiple different types of data storage media is described in commonly owned, U.S. Patent Application ______, (Attorney Docket number HSJ9-2007-0208-US1) filed concurrently herewith, which is incorporated by reference herein.
In the embodiment of
By storing a portion of a file in two or more different types of data storage media, file system 112 significantly increases the reliability of the data associated with that portion of the file, without having to store the entire file on two different data storage media.
In addition, storing a portion of a file in two different types of data storage media can also significantly decrease the read access time for that portion of the file. For example, Flash memory devices typically have fast random access times and slower sequential data transfer times. Hard disk drives typically have fast sequential data transfer times for accessing a sequential range of physical addresses and slower random access times. Thus, file system 112 can significantly decrease the read access time of data that is stored on both Flash memory and on a magnetic hard disk. Each time that host 101 requests data, the data are accessed from both the Flash memory and the hard disk. Whichever data storage medium is faster at accessing that data returns the data to the controller first. Then, the data received at the controller first are transferred to host 101.
Storing a portion of a file in two different types of data storage media can also significantly decrease the write time. For example, file system 112 can immediately write data associated with a portion of a file to the data storage medium that performs faster write operations. Then, file system 112 can copy the data associated with that portion of the file from the faster data storage medium to the other, slower data storage medium in the background, after the data has been written to the faster data storage medium. This technique ensures that the data is written onto at least one data storage medium soon after the write command is issued, while the data is copied to a second data storage medium later when host 101 is less busy.
According to yet another embodiment, data storage system 102 or 202 can store a portion of a file in three (or more) different types of data storage media.
According to some embodiments, a file system dynamically moves one or more portions of a file from one type of data storage medium to a different type of data storage medium if the intrinsic value of one or more portions of the file changes. For example, a file system can dynamically move a portion of a file from Flash memory to an optical disk or to a magnetic hard disk to increase the reliability of that portion of the file or to increase the sequential access time. As another example, a file system can move a portion of a file to a different type of data storage medium if the original data storage medium allocated to that portion of the file crashes or no longer provides a fast data access time for a particular application.
At step 402, file system 112 determines that the intrinsic value of one of the files has changed. Alternatively, file system 112 can change the intrinsic value of one of the files at step 402. At step 403, file system 112 moves the file from a first data storage medium to a second different data storage medium based on the new intrinsic value of that file. The first and the second data storage media are different types of data storage media.
According to another embodiment, file system 112 dynamically moves a portion of a file from one overlapping data storage medium to another. For example, file system 112 can move portion 301C of file 301 from data storage medium 105 to data storage medium 104. Portion 301C of file 301 then overlaps media 104 and 106.
The foregoing description of the exemplary embodiments of the present invention has been presented for the purposes of illustration and description. It is not intended to be exhaustive or to limit the present invention to the examples disclosed herein. A latitude of modification, various changes, and substitutions are intended in the present invention. In some instances, features of the present invention can be employed without a corresponding use of other features as set forth. Many modifications and variations are possible in light of the above teachings, without departing from the scope of the present invention.
For example, embodiments of the present invention may be implemented using one or a combination of hardware, software, and a computer-readable medium containing program instructions. Software implemented by embodiments of the present invention and results of the present invention may be stored on a computer-readable medium such as memory, hard disk drive, CD, DVD, or other media. Results of the present invention may be used for various purposes such as being executed or processed by a processor, being displayed to a user, transmitted in a signal over a network, etc.
