Uniprocessor Scheduling

Priority Queueing

• scheduler will always choose higher priority processes over lower priority ones
• use multiple ready queues to represent multiple leverls of priority
• lower priority may be starved

Multiprogramming

• nonpreemptive: once a process is in the running state, it continues until it terminates or gets blocked
• preemptive: current process may be interrupted and moved to ready state by OS. allows for better service since no one can monopolize the processor for too long
• cooperative

First-come-first-served

• Non-preemptive
• each process joins the ready queue
• when the current process finishes, the oldest one in the queue is picked to run next
• short process may have to wait a long time before execution
• favors CPU-bound processes

Round Robin (Time Slicing)

• Uses clock-based preemption
• Clock interrupt is generated at periodic intervals
• when interrupt occurs, the currently running process is put in the ready queue and next job is selected
• quantum size should be around the same size as the typical time of what they do between calls (e.g. UI)

Shortest process next

• non preemptive
• process with shortest expected processing time is selected next based on estimation
• shorter porcesses jump ahead of larger ones
• predicatability of longer processes is reduced
• if estimation isn't correct, the OS can abort the process
• possible starvation of larger processe
• limitation: knowing service time

Shortest Remaining Time

• preemptive version of SPN policy
• Must estimate processing time
• starvation of long processes still

Highest Response Ratio Next

• Choose next process with greatest ratio of time spent waiting to how much time the system will spend executing
• accounts for aging to prevent starvation

Feedback-based

penalize jobs that run longer
each time you are preemptive you go into a lower and lower priority queue

Fair Share

• schedules processes as groups so if something spawns a bunch of processes, it doesnt get better treatement in the scheduler