Scheduling With Inexact Job Sizes: The Merits of Shortest Processing Time First
It is well known that size-based scheduling policies, which take into account job size (i.e., the time it takes to run them), can perform very desirably in terms of both response time and fairness. Unfortunately, the requirement of knowing a priori the exact job size is a major obstacle which is fre...
Gespeichert in:
1. Verfasser: | |
---|---|
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | It is well known that size-based scheduling policies, which take into account
job size (i.e., the time it takes to run them), can perform very desirably in
terms of both response time and fairness. Unfortunately, the requirement of
knowing a priori the exact job size is a major obstacle which is frequently
insurmountable in practice. Often, it is possible to get a coarse estimation of
job size, but unfortunately analytical results with inexact job sizes are
challenging to obtain, and simulation-based studies show that several
size-based algorithm are severely impacted by job estimation errors. For
example, Shortest Remaining Processing Time (SRPT), which yields optimal mean
sojourn time when job sizes are known exactly, can drastically underperform
when it is fed inexact job sizes.
Some algorithms have been proposed to better handle size estimation errors,
but they are somewhat complex and this makes their analysis challenging. We
consider Shortest Processing Time (SPT), a simplification of SRPT that skips
the update of "remaining" job size and results in a preemptive algorithm that
simply schedules the job with the shortest estimated processing time. When job
size is inexact, SPT performs comparably to the best known algorithms in the
presence of errors, while being definitely simpler. In this work, SPT is
evaluated through simulation, showing near-optimal performance in many cases,
with the hope that its simplicity can open the way to analytical evaluation
even when inexact inputs are considered. |
---|---|
DOI: | 10.48550/arxiv.1907.04824 |