There are many works on cloud task scheduling that mainly address run-time efficient scheduling of tasks submitted to a cloud. The problem is much similar to the generic problem of scheduling processes in a generic operating systems. Average turnaround time is a system oriented metric that is optimized in process scheduling algorithms such as shortest-job-first or highest-response-ratio-next.

A point that is confusing for me is why scheduling in cloud service centers is treated so much differently than generic OSs ? What are cloud specific challenges that lead to evolutionary solutions instead of deterministic algorithms for generic OSs?

Similar questions and discussions