TY - GEN
T1 - Competitive analysis of task scheduling algorithms on a fault-prone machine and the impact of resource augmentation
AU - Fernández Anta, Antonio
AU - Georgiou, Chryssis
AU - Kowalski, Dariusz R.
AU - Zavou, Elli
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2015.
PY - 2015
Y1 - 2015
N2 - Reliable task execution on machines that are prone to unpredictable crashes and restarts is both important and challenging, but not much work exists on the analysis of such systems. We consider the online version of the problem, with tasks arriving over time at a single machine under worst-case assumptions. We analyze the fault-tolerant properties of four popular scheduling algorithms: Longest In System (LIS), Shortest In System (SIS), Largest Processing Time (LPT) and Shortest Processing Time (SPT). We use three metrics for the evaluation and comparison of their competitive performance, namely, completed load, pending load, and latency. We also investigate the effect of resource augmentation in their performance, by increasing the speed of the machine. Hence, we compare the behavior of the algorithms for different speed intervals and show that there is no clear winner with respect to all the three considered metrics. While SPT is the only algorithm that achieves competitiveness on completed load for small speed, LIS is the only one that achieves competitiveness on latency (for large enough speed).
AB - Reliable task execution on machines that are prone to unpredictable crashes and restarts is both important and challenging, but not much work exists on the analysis of such systems. We consider the online version of the problem, with tasks arriving over time at a single machine under worst-case assumptions. We analyze the fault-tolerant properties of four popular scheduling algorithms: Longest In System (LIS), Shortest In System (SIS), Largest Processing Time (LPT) and Shortest Processing Time (SPT). We use three metrics for the evaluation and comparison of their competitive performance, namely, completed load, pending load, and latency. We also investigate the effect of resource augmentation in their performance, by increasing the speed of the machine. Hence, we compare the behavior of the algorithms for different speed intervals and show that there is no clear winner with respect to all the three considered metrics. While SPT is the only algorithm that achieves competitiveness on completed load for small speed, LIS is the only one that achieves competitiveness on latency (for large enough speed).
KW - Competitive analysis
KW - Failures
KW - Online algorithms
KW - Resource augmentation
KW - Scheduling
KW - Task sizes
UR - http://www.scopus.com/inward/record.url?scp=84955259625&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84955259625&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-28448-4_1
DO - 10.1007/978-3-319-28448-4_1
M3 - Conference contribution
AN - SCOPUS:84955259625
SN - 9783319284477
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 1
EP - 16
BT - Adaptive Resource Management and Scheduling for Cloud Computing - 2nd International Workshop, ARMS-CC 2015 Held in Conjunction with ACM Symposium on Principles of Distributed Computing, PODC 2015 Donostia-San Sebastián, Spain, July 20, 2015, Revised Selected Papers
A2 - Pop, Florin
A2 - Potop-Butucaru, Maria
PB - Springer Verlag
T2 - 2nd International Workshop on Adaptive Resource Management and Scheduling for Cloud Computing, ARMS-CC 2015
Y2 - 20 July 2015 through 20 July 2015
ER -