It has been found that in a certain disk drive about 80% of all requests are for a small...

Question

It has been found that in a certain disk drive about 80% of all requests are for a small, fixed number of cylinders. Which of the scheduling algorithms will you recommend for this situation? Justify your choice.

Answer 1

For a situation where a large majority of requests (around 80%) are concentrated on a small, fixed number of cylinders in a disk drive, I would recommend using the SCAN (Elevator) scheduling algorithm.

The SCAN algorithm works by servicing requests in a sweeping motion across the disk, moving in one direction (from one end to the other) until it reaches the last request in that direction and then reversing its direction. This back-and-forth movement resembles the path of an elevator, hence the name.

Here's why the SCAN algorithm is a suitable choice for the given scenario:

1. Efficient for localized requests: The SCAN algorithm is efficient in handling localized requests because it moves the disk arm continuously across the entire disk. By doing so, it ensures that requests on the nearby cylinders are serviced quickly and effectively. Since 80% of the requests are concentrated on a small number of cylinders, SCAN can efficiently process these requests in a shorter time compared to other algorithms.
2. Avoids starvation: The SCAN algorithm guarantees that every request will eventually be serviced. By scanning the entire disk surface, it ensures fairness in servicing requests from different areas of the disk. This is important in situations where a fixed number of cylinders have a high volume of requests, as it prevents any particular request or region from being continually skipped or starved.
3. Predictable servicing time: SCAN offers predictable servicing time for requests. Since it follows a specific pattern of movement across the disk, the maximum time required to service a request can be estimated. This predictability helps in resource planning and system optimization.
4. Minimal head movement: The SCAN algorithm minimizes head movement by always sweeping the disk surface in a single direction until reaching the end, reducing the time wasted on seeking and rotational delays. This characteristic improves overall disk performance.

Although other algorithms like FCFS (First-Come, First-Served) or SSTF (Shortest Seek Time First) can also handle this scenario, SCAN provides better efficiency and fairness when the requests are localized to a specific area of the disk.