When you learn to quit, it
becomes a habit.
-Vince Lombardi
SCHEDULING: An Introduction
‘Schedule’
in literal
sense means plan for something to happen at a particular time, but in the terms
of operating system ‘Schedule’ or ‘Scheduling’ means to manage or
to plan the process such that it has minimum utilization of Central Processing
unit(CPU).
The
policies which are used for scheduling are called ‘Scheduling Policies’ (or
disciplines).
1.1
Workload Assumptions
Here
we will make some assumptions that will be implied as we move further.The
assumptions we make are:-
1)All
processes/workload run for same amount of time.
2)Each
Process start at the same time.
3)When
run each process runs to its completion.
4)Each
process uses CPU.
5)The
run time of each process is known.
1.2
Scheduling Metrices
A
‘Metric’ is something used to measure something.
Let
me introduce you to our first metric i.e. Turnaround time.
The
formula for Turnaround time is as follows:
T(turnaround)=
T(completion)-T(arrival)
where,
T(turnaround)= Turnaround time
T(completion)=
Completion Time
T(arrival)=
Arrival Time
Since,
as we stated in our assumption no. 2 i.e.:-
Each
process start at the same time, so,
T(arrival)=
0 seconds (for every process)
So,
the above equation becomes,
T(turnaround)=
T(completion)- T(arrival)
T(turnaround)=
T(completion) - 0
T(turnaround)=
T(completion)
This
equation will be applicable for further tutorials.
‘Turnaround
time’ and another term called ‘fairness’ are two main point of focus in
scheduling processes.
Since,
we already discussed ‘Turnaround time’, it’s time for our new term ‘Fairness’.
So,
what does ‘Fairness’ in terms of Scheduling means. Here, the term ‘Fairness’
means increasing performance by running one process/job to its fullest and
restricting others.
