Open Access
ARTICLE
QoS-Aware Energy-Efficient Task Scheduling on HPC Cloud Infrastructures Using Swarm-Intelligence Meta-Heuristics
Amit Chhabra1, *, Gurvinder Singh2, Karanjeet Singh Kahlon2
1 Department of Computer Engineering & Technology, Guru Nanak Dev University, Amritsar, 143005, India.
2 Department of Computer Science, Guru Nanak Dev University, Amritsar, 143005, India.
* Corresponding Author: Amit Chhabra. Email: .
Computers, Materials & Continua 2020, 64(2), 813-834. https://doi.org/10.32604/cmc.2020.010934
Received 08 April 2020; Accepted 27 April 2020; Issue published 10 June 2020
Abstract
Cloud computing infrastructure has been evolving as a cost-effective platform
for providing computational resources in the form of high-performance computing as a
service (HPCaaS) to users for executing HPC applications. However, the broader use of
the Cloud services, the rapid increase in the size, and the capacity of Cloud data centers
bring a remarkable rise in energy consumption leading to a significant rise in the system
provider expenses and carbon emissions in the environment. Besides this, users have
become more demanding in terms of Quality-of-service (QoS) expectations in terms of
execution time, budget cost, utilization, and makespan. This situation calls for the design
of task scheduling policy, which ensures efficient task sequencing and allocation of
computing resources to tasks to meet the trade-off between QoS promises and service
provider requirements. Moreover, the task scheduling in the Cloud is a prevalent NPHard problem. Motivated by these concerns, this paper introduces and implements a
QoS-aware Energy-Efficient Scheduling policy called as CSPSO, for scheduling tasks in
Cloud systems to reduce the energy consumption of cloud resources and minimize the
makespan of workload. The proposed multi-objective CSPSO policy hybridizes the
search qualities of two robust metaheuristics viz. cuckoo search (CS) and particle swarm
optimization (PSO) to overcome the slow convergence and lack of diversity of standard
CS algorithm. A fitness-aware resource allocation (FARA) heuristic was developed and
used by the proposed policy to allocate resources to tasks efficiently. A velocity update
mechanism for cuckoo individuals is designed and incorporated in the proposed CSPSO
policy. Further, the proposed scheduling policy has been implemented in the CloudSim
simulator and tested with real supercomputing workload traces. The comparative analysis
validated that the proposed scheduling policy can produce efficient schedules with better
performance over other well-known heuristics and meta-heuristics scheduling policies.
Keywords
Cite This Article
A. Chhabra, G. Singh and K. Singh Kahlon, "Qos-aware energy-efficient task scheduling on hpc cloud infrastructures using swarm-intelligence meta-heuristics,"
Computers, Materials & Continua, vol. 64, no.2, pp. 813–834, 2020. https://doi.org/10.32604/cmc.2020.010934
Citations