Operating system job scheduling and low-level scheduling algorithms >> computational Problems

As a computer professional students, operating system This course is the most serious, this is not, the midterm examination of a wave of questions, because usually do less problems, the test will inevitably disoriented,

Even though the difficult, hey, below I on the low-level scheduling algorithm (should be a very important test center, computational questions) some concepts to do a record, summed up, not quasi-final use on.

Some basic concepts and formulas for jobs (job):

Job Turnaround time = Job completion point-the job arrives at the system point in time

Take Ownership turnaround time = job turnaround time/Required CPU time (job processing time) = Response ratio

Yes, the weighted turnaround time is exactly the same as the response value, but the concept is different, and there is a conversion formula: response than =1+ job wait time/job processing time

Scheduling algorithm:

1. First come first served (FCFS)

The simplest algorithm, in order to select the job according to queue sequence, first enter the system of the job will be selected into the memory priority

2. Shortest job Priority algorithm (SJF)

Pick the minimum required CPU time in the queue to perform the job first. However, the implementation of the SJF algorithm needs to know the job run time beforehand, otherwise there is no basis for scheduling.

3. Shortest remaining time Priority algorithm (SRTF)

A stripped-down scheduling algorithm that assumes that a process/thread is currently running, and that if a new process/thread is moved into the ready queue, the preemptive shortest job priority algorithm will forcibly deprive the current performer of control if the CPU run time required is shorter than the remaining CPU time required by the current running process/thread. Schedule a new process/thread execution, which is called the remaining shortest remaining time priority algorithm.

4. Highest response ratio priority algorithm (HRRF)

Assuming that there are multiple jobs in a job queue, first perform the job that first entered the system

And then calculate the response ratio of the other remaining jobs, select the one that performs the highest response

Then, to calculate the response ratio of the remaining other jobs, select the response than the highest execution ...

Loop several times until the job in the queue finishes executing