Homework for Chapter 7 of 《Operating Systems: Three Easy Pieces》: Scheduling Basics

Concepts:

the “response time” (the time a job spends waiting after arrival before first running);
the “turnaround time” (the time it took to complete the job since first arrival);
the total “wait time” (any time spent ready but not running).

1&2:
I will simply combine both of 2 questions and draw some conclusion:

root@C202511211157657:~\# ./scheduler.py -p FIFO -j 3 -s 1
ARG policy FIFO
ARG jobs 3
ARG maxlen 10
ARG seed 1
Here is the job list, with the run time of each job:
  Job 0 ( length = 2 )
  Job 1 ( length = 9 )
  Job 2 ( length = 8 )
Compute the turnaround time, response time, and wait time for each job.
When you are done, run this program again, with the same arguments,
but with -c, which will thus provide you with the answers. You can use
-s <somenumber> or your own job list (-l 10,15,20 for example)
to generate different problems for yourself.

Its policy is FIFO, so we could add up the execution times:

Actually: Correct.

Notice that the wait time and the response time is always the same as job won’t be interrupted since started.

root@C202511211157657:~# ./scheduler.py -p SJF -j 3 -s 1
ARG policy SJF
ARG jobs 3
ARG maxlen 10
ARG seed 1
Here is the job list, with the run time of each job:
  Job 0 ( length = 2 )
  Job 1 ( length = 9 )
  Job 2 ( length = 8 )
Compute the turnaround time, response time, and wait time for each job.
When you are done, run this program again, with the same arguments,
but with -c, which will thus provide you with the answers. You can use
-s <somenumber> or your own job list (-l 10,15,20 for example)
to generate different problems for yourself.

SJF need us to do some extra sorting, but the calculation is almost the same:

Actually: Correct.

Slightly improved.

root@C202511211157657:~# ./scheduler.py -p RR -l 100,200,300 -q 1
ARG policy RR
ARG jlist 100,200,300

Here is the job list, with the run time of each job:
  Job 0 ( length = 100.0 )
  Job 1 ( length = 200.0 )
  Job 2 ( length = 300.0 )


Compute the turnaround time, response time, and wait time for each job.
When you are done, run this program again, with the same arguments,
but with -c, which will thus provide you with the answers. You can use
-s <somenumber> or your own job list (-l 10,15,20 for example)
to generate different problems for yourself.

Turnaround time is pretty difficult to calculate… The other two are simple.

Actually: Correct.

4. When FIFO Queue is ordered.

5. When there is only one (or less) task of which length is more than the quantum length.

6. Grows longer.

7. Grows longer, and decade to FIFO.

Worst response time = (max_length + total_length) * N / 2.

So it would be affcted linearly by the job length and the numbr of jobs.