A Comprehensive Study of Load Balancing Approaches in the Cloud Computing Environment and a Novel Fault Tolerance Approach

Abstract

The past few years have witnessed the emergence of a novel paradigm called cloud computing. CC aims to provide computation and resources over the internet via dynamic provisioning of services. There are several challenges and issues associated with implementation of CC. This research paper deliberates on one of CC main problems i.e. load balancing (LB). The goal of LB is equilibrating the computation on the cloud servers such that no host is under/ overloaded. Several LB algorithms have been implemented in literature to provide effective administration and satisfying customer requests for appropriate cloud nodes, to improve the overall efficiency of cloud services, and to provide the end user with more satisfaction. An efficient LB algorithm improves efficiency and asset's usage through effectively spreading the workload across the system's different nodes. This review research paper objective is to present critical study of existing techniques of LB, to discuss various LB parameters i.e. throughput, performance, migration time, response time, overhead, resource usage, scalability, fault tolerance, power savings, etc. The research paper also discusses the problems of LB in the CC environment and identifies the need for a novel LB algorithm that employs FT metrics. It has been found that traditional LB algorithms are not good enough and they do not consider FT efficiency metrics for their operation. Hence, the research paper identifies the need for FT efficiency metric in LB algorithms which is one of the main concerns in cloud environments. A novel algorithm that employs FT in LB is therefore proposed.

Fault-Tolerant Load Balancing in Cloud Computing: A Systematic Literature Review

“…• Resource utilization: It specifies that what degree of VMs uses the tools. In fact, it determines a part of accessible services among the total available resources [50]. • Makespan: It is defined as the time taken to process a set of tasks for its complete execution [51].…”

Simulation and performance assessment of a modified throttled load balancing algorithm in cloud computing environment

“…In this respect, several loadbalancers have been proposed considering the aforementioned aspects [10]- [22]. Among the algorithms, round robin (RR), active monitoring load balancer (AMLB), throttled load balancer (TLB) are common load balancing algorithms in large-scale cloud environment using cloud simulator [23]- [29].…”

Multilevel controller‐assisted intrinsically modified ant colony optimization heuristic‐based load‐balancing model for mega cloud infrastructures

“…Although the use of prediction concept can help avoiding false‐positive scheduling, not significant efforts are made so far. Furthermore, for SLA‐centric cloud service provision, there is the need of network‐sensitive allocation even under constrained SLA conditions 5,12,17,18 . Unlike classical bin packing‐oriented VM consolidation concepts, to ensure QoS demands, the migration controller requires considering other performance aspects as well, especially under dynamic cloud condition 19–24 .…”

“…To alleviate above‐stated convexity problem or NP‐hard problem, heuristic algorithms or evolutionary computing methods have gained widespread attention 28,29 . In our targeted VM migration of load‐balancing task, heuristic method(s) can be applied to identify the best suitable host or PM to attach the VM or task from the overloaded host 17,18,27–31 . To be noted that in the classical bin‐packing methods, algorithms 32 often use random host selection and therefore do not care the ongoing tasks and its continuity issues, resource demands, and so on 19–24,27 .…”