Super block (file system)
The super block (British: Superblock, Super block) is meta data to manage the logic partition of the file system [1]; [2]. It is used by the Unix system OS's mainly [1]. The super block becomes the important clue to recognize the file system of the block device at the time of system start as follows.
Table of contents
Linux
An example includes Linux. The file system of Linux has the hierarchical block structure including a super block, inode block, "a directory management block" [3]; [4]. As for these, basic information of the file system including the size of the total number of inode which the whole file system has and 残数 and the file system of inode is stored. In addition, the virtual file system (VFS) maintains the block information mentioned above separately from the file system which a user uses [5]. At first inode information of the root directory ("/") is necessary for the file reading from a file system mount (mount command) and file system of () at the time of Linux system start, but the information is written to the super block. I can discover inode of the finally objective file and become accessible if I call it sequentially because there is inode block in a directory in one as for the root directory following directories [6]. In the case of the system which has software RAID toward a storage, the information about the multi-device is necessary like the above for a mount of the file system. The information is written in at the super block of the block device to use for RAID [7]. In particular, such a super block called the Percy stent supermarket block (British: Persistent superblock, a permanent supermarket block) [7] [8] [9]. When I initialize RAID array, this information is removed by the super block of all array constitution disks.
When file system is damaged; of the block usually super using fsck command the information of file system in itself is compared with the information (inode and the directory), and or restoration is hand-operated automatically, and can restore it so that consistency is kept [10]. However, as for this restoration processing, time to take grows big when disk space becomes big. Therefore, in late years a journal is used for restoration of the file system. ジャーナリングファイルシステム is structure speeding up restoration handling of file system and, in the case of the update of the meta data such as super blocks, updates the log of the journal beforehand when I easily speak it. It is not in this way necessary to compare the true data with the direct super block at the time of file system damage [2]. However, I cannot evade meta data corruption in ジャーナリングファイルシステム either. A super block was destroyed or can let mkfs () [explanatory note 1] and fsck[ explanatory note 2] command restore in combination when I maintain spare Sioux pub lock by any chance when it becomes impossible of reading. I can make it immediately when a super block does not exist.
The OS always updates a super block of the file system (when, thus, some kind of disorders occur, a super block may be damaged). When the data buffer of the super block which I updated on the memory exists, I can write in it by sync () command immediately [11].
Other
With the thing which the implementation is different even as for the pro-other Unix OS, but constitutes a supermarket block in file system by a similar policy [12] [13].
Footnote
Explanatory note
- In the case of ^ ext3 and ext4, I understand a block address and the block size that there is the backup of the super block when I carry out "-n" option and file system in an argument. Noriyuki Kitaura (April 21, 2005). "To restore a broken partition". At mark アイティ. www.atmarkit.co.jp. August 10, 2011 reading.
- A check of the file system is enabled after reading the Sioux pub lock when I appoint the block address of the super block with "-b" option in ^ ext3 and ext4. Noriyuki Kitaura (April 21, 2005). "To restore a broken partition". At mark アイティ. www.atmarkit.co.jp. August 10, 2011 reading.
Source
- ^ a b "Structure of the file system". stwww.eng.kagawa-u.ac.jp. August 10, 2011 reading.
- ^ a b Sugaya green (July 29, 2003). ""The information" that Linux file system technology commentary - third ジャーナリングファイルシステム protects". At mark アイティ. www.atmarkit.co.jp. August 11, 2011 reading.
- ^ M. Tim Jones (October 30, 2007). "Thorough investigation into Linux file system". IBM. www.ibm.com. August 10, 2011 reading.
- ^ Shiyu Mori (March 2, 2010). "Data management (1/2) of the splendid file system". At mark アイティ. monoist.atmarkit.co.jp. August 10, 2011 reading.
- ^ David A Rusling (January 25, 1999). "The Linux Kernel: file system". David A Rusling, Linux JF Project. linuxjf.sourceforge.jp. August 10, 2011 reading.
- ^ Greg O'Keefe (2000). "From Power To Bash Prompt which improves". Greg O'Keefe, 1,000 Yuji Dan. linuxjf.sourceforge.jp. August 10, 2011 reading.
- ^ a b "Boot time assembly of RAID arrays." git.kernel.org. August 10, 2011 reading.
- ^ "md: nonpersistent superblock ...". OSS Message Pedia. ossmpedia.org. August 11, 2011 reading.
- ^ JakobØstergaard. "4.7 The Persistent Superblock". JakobØstergaard, Takashi Kasai. linuxjf.sourceforge.jp. August 11, 2011 reading.
- ^ Sugaya green (June 24, 2003). "Linux file system technology commentary - second block algorithm and B-Tree algorithm". At mark アイティ. www.atmarkit.co.jp. August 11, 2011 reading.
- ^ "Manpage of SYNC". linuxjm.sourceforge.jp. August 10, 2011 reading.
- ^ Solaris, UFSA の example. "System management (Vol. 1) - supermarket block of Solaris". Oracle. download.oracle.com. August 11, 2011 reading.
- ^ AIX, JFSA の example. "Placement of the JFS file system". IBM. publib.boulder.ibm.com. August 11, 2011 reading.
Allied item
Outside link
This article is taken from the Japanese Wikipedia Super block (file system)
This article is distributed by cc-by-sa or GFDL license in accordance with the provisions of Wikipedia.
In addition, Tranpedia is simply not responsible for any show is only by translating the writings of foreign licenses that are compatible with CC-BY-SA license information.
0 개의 댓글:
댓글 쓰기