Describe the following disk head scheduling algorithms: FIFO, SSTF, SCAN, C?SCAN?

What will be an ideal response?

FIFO: first?in?first?out scheduling processes requests in the order in which they are received (it’s a sort of nonalgorithm
in the sense that it is the default action). It is inefficient because tracks are accessed at random and
this may involve a lot of head movement. It is also described as fair because all processes are treated equally.

SSTF: shortest service time scheduling first selects the next track to seek as the one that is closest (i.e.,
shortest seek time). This is fast because it avoids large swings of the head by dealing with clusters of tracks. It
is unfair because a process may have widely distributed tracks and be left waiting while clustered tracks from
another process are serviced first.

SCAN: SCAN reduces the number of changes in head direction by sweeping the head across the disk scanning
tracks and accessing them as the scan is made. To do this, the requests must be buffered in order of track
number. When the head reaches the edge of the disk, the scan is reversed and continues in the reverse
direction. This is basically like the action of an elevator that stops at floors on the way up where people want
to go up, then at floors on the way down where people want to go down.

C_SCAN: This is virtually the same as SCAN except that the scan takes place in one direction only. At the end
of a scan, the head moves back to begin another scan.