Linux File Systems

The Linux file system is the basic of each installation. Our new series introduces the administration of the file system and explains the first part of the main file system types that Linux offers. Our new Linux-series describes the various facets of the administration of the file system. With the exception of the first two sections of the chapter is directed specifically at Linux professionals. The eight-part series addressed the following issues:
File system types: This section provides an overview of features supported by Linux file system types.

• Management of the file system: In this section you will learn how to Linux and Windows partitions, floppy disks, data CDs and DVDs, USB memory sticks, etc.-can-use. The section also gives tips on what to do when a hard drive partition proves to be too small.
• Disk partitioning: Partitioning your hard drive is a central part of the installation of Linux. Sometimes it is also in the operation of Linux need to add a new partition.
• File systems: ext3 is the main Linux file system. It is backward compatible with its predecessor, ext2, but contains additional journaling. reiserfs is an alternative to ext3. The file system also supports journaling and is optimized especially for dealing with small files.
• RAID : RAID (Redundant Array of Inexpensive / Independent Disks) Link to the partitions of multiple hard drives together to fashion a more reliable and / or faster overall system to achieve this. This section will briefly discuss the basics of RAID, and then describes the creation of a RAID-0 system (striping).
• LVM: The logical volume manager (LVM short) allows flexible management of partitions. With LVM you can combine partitions such as multiple hard drives into a virtual partition, resize partitions on the fly.
• DMA mode for IDE hard drives: the so-called bus-master DMA mode can speed up access to IDE hard drives considerably. For compatibility reasons, this mode is often not enabled by default.

Using the Linux file system file system types are meant, in which stores a running Linux system has its own files and directories. In everyday operation, you will not notice with which file system you use. Elementary commands such as ls or cp, the management of access rights, etc. - all this works regardless of the file system. (Which or what file system type you are currently using, you can moreover easily identify with the command df-T). The file systems differ in characteristics that mainly for advanced users or for server -interest are used: speed in dealing with very large or very many small files, CPU utilization, journal function (behavior after a crash), Quota function (the ability to limit the maximum memory consumption per user), NFS-compatibility, administrative overhead, support for additional access rights (ACL), etc. In this respect, so there is no file system, which is simply the best - the rating depends on the purpose.

In the past, appeared repeatedly on the question of how large files can be up to. The answer depends on which kernel, which CPU architecture that use glibc library, and which volume you. Current distributions support throughout the LFS extensions (Large File System Support.) In the glibc library. Therefore cases, the maximum file size limit in most (2 ^ 63 bytes, which is more than on hard drives can ever be saved).

The following are some Linux File Systems :

• ext2: ext2 (extended file system version 2) was for many years the dominant Linux file system. It is extremely stable and secure. The only one of these file systems it does not support so-called journaling. After a crash, therefore, must be a time-consuming review of the entire file system are conducted.
• ext3: Since 2002, the successor to ext3 ext2 has taken it upon. It is this highly compatible, but supports journaling and from kernel 2.6 and ACLs.
• reiserfs: The Reiser file system, the name of its initiator Hans Reiser is derived, was the first journaling file system functions, the input into the Linux kernel managed the (version 2.4.1). Compared to ext2 or ext3, the Reiser file system has a higher speed especially when dealing with many small files.
• xfs: xfs file system has long been on the SGI workstations under IRIX operating system as in use. It is particularly useful for managing very large databases (up into the TB range). It supports quotas and extended attributes (ACLs). The file system can be increased during operation (without umount).
• jfs: jfs stands for journaled file system was originally developed by IBM and later ported to Linux.

Journaling means in its simplest form, the beginning and end of every operation in a special file to be logged. Thanks to the Protocol can be checked later to see if a particular file operation was completed. If this is not the case, then the operation can be undone. In the database world this is called transactions. In advanced systems, journaling is also possible, the actual changes to the files in the Journal log. This slows down the normal operation, there are more opportunities for future reconstruction. Now, if a file operation such as a power failure can not be completed, these activities under the Protocol. If only a simple journaling was active, although the changes are lost (ie you expect miracles from the journaling function), the current state of the file is available but not usually available.