Shortest job next (SJN), also known as shortest job first (SJF) or shortest process next (SPN), is a scheduling policy that selects for execution the waiting process with the smallest execution time.[1] SJN is a non-preemptive algorithm. Shortest remaining time is a preemptive variant of SJN.
Shortest job next is advantageous because of its simplicity and because it minimizes the average amount of time each process has to wait until its execution is complete. However, it has the potential for process starvation for processes which will require a long time to complete if short processes are continually added. Highest response ratio next is similar but provides a solution to this problem using a technique called aging.[2]
Another disadvantage of using shortest job next is that the total execution time of a job must be known before execution. While it is impossible to predict execution time perfectly, several methods can be used to estimate it, such as a weighted average of previous execution times.[3] Multilevel feedback queue can also be used to approximate SJN without the need for the total execution time oracle.[1]
Shortest job next can be effectively used with interactive processes which generally follow a pattern of alternating between waiting for a command and executing it. If the execution burst of a process is regarded as a separate "job", the past behaviour can indicate which process to run next, based on an estimate of its running time.
Shortest job next is used in specialized environments where accurate estimates of running time are available.
- ^ a b Arpaci-Dusseau, Remzi H.; Arpaci-Dusseau, Andrea C. (2014), Operating Systems: Three Easy Pieces [Chapter Scheduling Introduction] (PDF), Arpaci-Dusseau Books
- ^ Tanenbaum, A. S. (2008). Modern Operating Systems (3rd ed.). Pearson Education, Inc. p. 156. ISBN 978-0-13-600663-3.
- ^ Silberschatz, A.; Galvin, P.B.; Gagne, G. (2005). Operating Systems Concepts (7th ed.). Wiley. p. 161. ISBN 0-471-69466-5.