| We know that we must have at least one filesystem (the root filesystem), but why do we bother with any others? There are a number of reasons for splitting a disk into filesystems and some of them are historical. For example, the System V filesystem type did not scale well, so as the size increased the performance fell. Consequently, it was advantageous to have many smaller filesystems rather than a few bigger ones. Also, a filesystem has a maximum size and a maximum number of files that it can hold (dependent on the filesystem type), and the limits increased as new filesystem types were developed. As long as our disk was small enough, we could choose to have just one large filesystem, or if we had many disks we could choose one filesystem per disk. This would greatly simplify the set up of the system and it should reduce the problems of filesystems filling up. This may be a sensible way of configuring a system, but it also may not be. It all depends on the way the system will be used. If the system has many applications installed, then it is sensible to keep these in separate filesystems so they don't interfere with each other. If one application goes berserk and starts filling up the disk, it will stop when its own filesystem is full and won't cause any problems to other applications. Also, if an application is to be taken offline for any reason, the filesystem could be physically unmountedwhich would prevent anybody from playing with it. It is also often desirable to keep the application code and data on separate disks (and even separate disk controllers) to prevent performance bottlenecks; this can only be achieved by using a combination of separate filesystems on separate disks. |
