6. Scheduling

The purpose of scheduling

1. Ordering resource assignment
2. Do prediction of the system behavior under anticipated loads.
   In running time: to permit the new request of the loading-balancing
   In compile time: to confirm the feasibility of the system or predict resource needs.

Criteria in Performance

Process /user perspective: Waiting time, Response time（从用户提交请求开始，直达系统首次产生响应为止的时间间隔）, Turnaround Time(完成时间 - 提交时间).
System perspective: Throughput， Utilization.

Method in scheduling

1. First come first served

2. Round Robin
   A round-robin scheduler generally employs time-sharing, giving each job a time slot or quantum[4](its allowance of CPU time), and interrupting the job if it is not completed by then.

3. Feedback with 2^i pre-empty intervals
   ---- When processes can be readily categorized, then multiple separate queues can be established, lower priority more 2^i unit time.
   ---- Scheduling must also be done between queues, that is scheduling one queue to get time relative to other queues.
   ---- Jobs may be moved from one queue to another for a variety of reasons
   ---- Scheduling is the most flexible, because it can be tuned for any situation

4. Shortest job first
   when the computing time is know and switch is expensive then this method is good.

5. Highest Response Ration First(HRRF)
   Combine the "Short job first" and "First come first serve".

6. Shortest remaining time first
   The difference between 6 and 4 is that this one has ability to pre-emption the current executing job.

Predictable scheduling

guarantee task:

1. the deadline
2. reaction time
3. data flow level
4. delivery time
5. the bound of variables in result.