Previously, we discussed schedulers in terms of turnaround and response times. This time we want to look a scheduler that dictates that each job gets a specified percentage of CPU time. This is why it is called a fair$-$share scheduler.

The operating system repeatedly holds a lottery to determine which job is run next. Processes that should be run with greater frequency should have a better chance of winning the lottery.

Instead of optimizing for turnaround or response time, a scheduler can seek to guarantee that each job gets a certain percentage of CPU time.

The idea is to hold a lottery every now and then to choose which process should run next. Processes that should run more often should have greater chances to win.

How does the lottery system work?