Shortest Job First Scheduling

<<Previous

Next >>

Shortest Job First Scheduling or SJF allocates the CPU to the process which has the smallest CPU-bursts time.

• SJF is also called Shortest Job Next scheduling shortly referred to as SJN. Since Job can be called as a process, SJF is also referred to as Shortest Process Next or SPN scheduling

• Suppose, if the CPU bursts of two processes are the same, then we use FCFS scheduling to break the tie.

• SJF is used frequently in long term scheduling but it cannot be implemented in short-term scheduling.

• SJF can either be a preemptive or nonpreemptive.

Advantages of SJF

Pros of using SJF:

• Shortest jobs execute first and do not wait for time consuming jobs.
• SJF helps achieve lesser average waiting time
• More jobs complete faster since long running jobs are executed later

• Helps achieve lesser turnaround time

• SJF is simple and hence reduces overhead
• SJF is most suitable for environments where estimated time for each job is available
• SJF most sutiable for interactive processes which wait for a trigger and then execute

Limitations of SJF

Disadvantages/drawbacks/cons of using SJF:

• Jobs or processes that need more time will wait and result in Job starvation or process starvation

Preemptive SJF scheduling

Preemptive SJF scheduling is sometime called Shortest remaining time first Scheduling.

How does it work?

Preemptive means a Job can be stopped midway to give CPU time to another Job. In Preemptive SJF, the job which has minimum remaining time will be executed next

SJF Example 1 - For Processes That Arrive At The Same Time

For example, consider the set of processes that arrive at time 0 and CPU burst time given in millisecond:

SJF Example 1 - Gantt Chart

In the above diagram: 3,7 and 27 are finish times

SJF Example 1 - Waiting Time and Turnaround time

SJF Example 2 - For Processes That Arrive At Different Times

Let us say, the processes arrive in the different arrival time, and say, the process is served by using Preemptive SJF Scheduling.

Now, based on the processes which are in queue at the start:

• Process which takes the least time will be executed first.
• Once first process is progressing, other processes are checked.
• Algorithm will calculate the remaining time for all the processes.
• Either the current process or any other process that arrived, which has the least remaining time will execute next.
• In a similar way, as more processes arrive, current process is paused and process with the least remaining time will be executed.

SJF Example 2 - Gantt Chart

SJF Example 2 -Waiting Time and Turnaround time

<<Previous

Next >>

Wait and Turnaround time example- For Processes That Arrive At Different Times