Shortest seek first

Shortest seek first, also called shortest seek time first, is a secondary storage scheduling algorithm that decides how the read and write head of a disk drive moves to serve requests. The operating system queues the incoming requests and always serves next the one that is closest to the cylinder and sector currently under the head, so a priority queue is ordered by the distance the head must travel. This is quicker than serving requests in the order they arrive, because less actuator arm movement is wasted on seeks, and the source calls the method an improvement over first come, first served. The cost is starvation: when many requests cluster in one region of the disk, requests that are physically far away may wait a long time. It targets magnetic hard disk drives, where seek time and rotational latency dominate the cost of an access, and it is a special case of the shortest job next scheduling algorithm.