Tuesday, May 20, 2014

Basic Linux Filesystems tutorial : ext2,ext3,ext4,JFS & XFS

The original Linux system used a simple filesystem that mimicked the functionality of the Unix filesystem. In this tutorial we will discuss basic file system used in Linux.

The ext Filesystem

The original filesystem introduced with the Linux operating system is called the extended filesystem (or just ext for short). It provides a basic Unix- like filesystem for Linux, using virtual directories to handle physical devices, and storing datain fixed-length blocks on the physical devices.

The ext filesystem uses a system called inodes to track information about the files stored in the virtual directory. The inode system creates a separate table on each physical device, called the inode table , to store the file information. Each stored file in the virtual directory has an entry in  the inode table. The extended part of the name comes from the additional data that it tracks on each file, which consists of:

  • The file name
  • The file size
  • The owner of the file
  • The group the file belongs to
  • Access permissions for the file
  • Pointers to each disk block that contains data from the file

Linux references each inode in the inode table using a unique number (called the inode number ), assigned by the filesystem as data files are created. The filesystem uses the inode number to identify the file rather than having to use the full file name and path.

The ext2 Filesystem

The original ext filesystem had quite a few limitations, such as limiting files to only 2GB in size. Not too long after Linux was first introduced, the ext filesystem was upgraded to create the second extended filesystem, called ext2 .
As you can guess, the ext2 filesystem is an expansion of the basic abilities of the ext filesystem, but maintains the same structure. The ext2  filesystem expands the inode table format to track additional information about each file on the system.

The ext2 inode table adds the created, modified, and last accessed time values for files to help system administrators track file access on the  system. The ext2 filesystem also increases the maximum file size allowed to 2TB (then in later versions of ext2, that was increased to 32TB) to help  accommodate large files commonly found in database servers.
In addition to expanding the inode table, the ext2 filesystem also changed the way in which files are stored in the data blocks. A common  problem with the ext filesystem was that as a file is written to the physical device, the blocks used to store the data tend to be scattered throughout the device (called fragmentation ). Fragmentation of data blocks can reduce the performance of the filesystem, as it takes longer to search the storage device to access all of the blocks for a specific file.

The ext2 filesystem helps reduce fragmentation by allocating disk blocks in groups when you save a file. By grouping the data blocks for a file, the filesystem doesn't have to search all over the physical device for the data blocks to read the file. The ext2 filesystem was the default filesystem used in Linux distributions for many years, but it, too, had its limitations. The inode table, while a  nice feature that allows the filesystem to track additional information about files, can cause problems that can be fatal to the system. Each time the filesystem stores or updates a file, it has to modify the inode table with the new information. The problem is that this isn't always a fluid action.

If something should happen to the computer system between the file being stored and the inode table being updated, the two would become out  of sync. The ext2 filesystem is notorious for easily becoming corrupted due to system crashes and power outages. Even if the file data is stored just  fine on the physical device, if the inode table entry wasn't completed, the ext2 filesystem wouldn't even know that the file existed!  It wasn't long before developers were exploring a different avenue of Linux filesystems. 

Journaling Filesystems

Journaling filesystems provide a new level of safety to the Linux system. Instead of writing data directly to the storage device and then updating the  inode table, journaling filesystems write file changes into a temporary file (called the journal ) first. After data is successfully written to the storage device and the inode table, the journal entry is deleted.

If the system should crash or suffer a power outage before the data can be written to the storage device, the journaling filesystem just reads through the journal file and processes any uncommitted data left over.

There are three different methods of journaling commonly used in Linux, each with different levels of protection. These are shown in  below Table.

Journaling Filesystem Methods

Data modeBoth inode and file data are journaled. Low risk of losing data, but poor performance.
Ordered modeOnly inode data written to the journal, but not removed until file data is successfully written. Good compromise between performance and safety.
Writeback modeOnly inode data written to the journal, no control over when the file data is written. Higher risk of losing data, but still better than not using journaling.


The data mode journaling method is by far the safest for protecting data, but it is also the slowest. All of the data written to a storage device must  be written twice, once to the journal, then again to the actual storage device. This can cause poor performance, especially for systems that do a lot  of data writing.  Over the years, a few different journaling filesystems have appeared in Linux. The following sections describe the popular Linux journaling  filesystems available.

The Extended Linux Journaling Filesystems

The same group that developed the ext and ext2 filesystems as part of the Linux project also created journaling versions of the filesystems. These journaling filesystems are compatible with the ext2 filesystem, and it's easy to convert back and forth between them. There are currently two  separate journaling filesystems based on the ext2 filesystem.


The ext3 Filesystem

The ext3 filesystem was added to the Linux kernel in 2001, and up until recently was the default filesystem used by just about all Linux distributions. It uses the same inode table structure as the ext2 filesystem, but adds a journal file to each storage device to journal the data written to the storage device.
By default, the ext3 filesystem uses the ordered mode method of journaling, only writing the inode information to the journal file, but not removing it until the data blocks have been successfully written to the storage device. You can change the journaling method used in the ext3 filesystem to either data or writeback modes with a simple command line option when creating the filesystem.

While the ext3 filesystem added basic journaling to the Linux filesystem, there were still a few things it lacked. For example, the ext3 filesystem doesn't provide any recovery from accidental deletion of files, there's no built-in data compressionavailable (although there is a patch that can be  installed separately that provides this feature), and the ext3 filesystem doesn't support encrypting files. For those reasons developers in the Linux project choose to continue work on improving the ext3 filesystem.

The ext4 Filesystem

The result of expanding the ext3 filesystem was (as you probably guessed) the ext4 filesystem. The ext4 filesystem was officially supported in the  Linux kernel in 2008, and is now the default filesystem used in most popular Linux distributions, such as Fedora and Ubuntu.

In addition to support compression and encryption, the ext4 filesystem also supports a feature called extents .Extents allocate space on a storage device in blocks, and only store the starting block location in the inode table. This helps save space in the inode table by not having to list all of the data blocks used to store data from the file.

The ext4 filesystem also incorporates block preallocation . If you want to reserve space on a storage device for a file that you know will grow in size, with the ext4 filesystem it's possible to allocate all of the expected blocks for the file, not just the blocks that physically exist. The ext4 filesystem fills in the reserved data blocks with zeroes, and knows not to allocate them for any other file.

The Reiser Filesystem

In 2001, Hans Reiser created the first journaling filesystem for Linux, called ReiserFS . The ReiserFS filesystem only supports writeback journaling mode, writing only the inode table data to the journal file. Because it writes only the inode table data to the journal, the ReiserFS filesystem is one of the fastest journaling filesystems in Linux.

Two interesting features incorporated into the ReiserFS filesystem are that you can resize an existing filesystem while it's still active, and that it uses a technique called tailpacking , which stuffs data from one file into empty space in a data block from another file. The active filesystem resizing feature is great if you have to expand an already created filesystem to accommodate more data.

The Journaled Filesystem(JFS)

Possibly one of the oldest journaling filesystems around, the Journaled File System (JFS) was developed by IBM in 1990 for its AIX flavor of Unix. However, it wasn't until its second version that it was ported to the Linux environment.

Note  :The official IBM name of the second version of the JFS filesystem is JFS2, but most Linux systems refer to it as just JFS.

The JFS filesystem uses the ordered journaling method, storing only the inode table data in the journal, but not removing it until the actual file data is written to the storage device. This method is a compromise between the speed of the ReiserFS and the integrity of the data mode journaling method.

The JFS filesystem uses extent-based file allocation, allocating a group of blocks for each file written to the storage device. This method provides for less fragmentation on the storage device. Outside of the IBM Linux offerings, the JFS filesystem isn't popularly used, but you may run into it in your Linux journey.

The XFS Filesystem

The XFS journaling filesystem is yet another filesystem originally created for a commercial Unix system that made its way into the Linux world. Silicon Graphics Incorporated (SGI) originally created XFS in 1994 for its commercial IRIX Unix system. It was released to the Linux environment for common use in 2002.

The XFS filesystem uses the writeback mode of journaling, which provides high performance but does introduce an amount of risk because the actual data isn't stored in the journal file. The XFS filesystem also allows online resizing of the filesystem, similar to the ReiserFS filesystem, except XFS filesystems can only be expanded and not shrunk.