7 Mounting file systems

DOS, and at least earlier versions of Windows, identify file systems as “drive letters”. In other words, c: is used to represent the primary partition of a hard disk drive. As additional file systems (in additional partitions and volumes) are used, other letters are used.

Linux and all unix-like operating systems do not do this. Instead, a file system can “mount” to a mount point of a parent (container) file system. The convention is that additional file systems (from external hard disk drives, thumb drives and etc.) are mounted to directories in the /media directory. However, this is merely a convention of Debian systems.

This flexible method is a direct result of the requirement that all file systems must look like VFS. The first inode of a file system can be easily “grafted” as the inode of a directory that belongs to a parent file system.

Note that the grafting is volatile. This means that upon reboot, Linux does not automatically remount file systems to their mount points. If automatic mounting is desired, then a file system along with its mount point can be specified in a “file system table” file, which is usually /etc/fstab.

The mount command is used to mount additional file systems. It is also used (by itself) to list all the mounted file systems.