Load balancing in cloud computing – A hierarchical taxonomical classification

Afzal, Shahbaz; Kavitha, G.

doi:10.1186/s13677-019-0146-7

Journal of Cloud Computing

Advances, Systems and Applications

Table 2 Reactive approaches of cloud load balancing in existing literature

From: Load balancing in cloud computing – A hierarchical taxonomical classification

Reference	Algorithm Used	State of Algorithm	Trait Used	Type of Load Balancing	Technique Involved	Algorithm Complexity	Advantages	Disadvantages
[48]	Conventional non classical Algorithm	Dynamic	Task Scheduling	Task LB	Non Classical, Deterministic	Not Specified	Better makespan	Task deadline not considered
							Better resource utilization	SLV not considered
							Less waiting time	Less fault tolerant
							Less execution time	Less energy efficient
[39]	Classical and Linear Programming	Dynamic	Task Scheduling	Task LB	Optimization (Linear programming Based)	Not Specified	Better makespan	Reduced quality of service
[39]	Classical and Linear Programming	Dynamic	Task Scheduling	Task LB	Optimization (Linear programming Based)	Not Specified	Better resource utilization	Reduced quality of service
[49]	GA and Min-Min	Hybrid	Task Scheduling	Task LB	Heuristic (Evolutionary)	O(m) and O (mn)	Better scalability	Less resource utilization
							Less response time	High SLV
							Small execution cost	Less degree of balance
[50]	BFO+ Lamarack Evolutionary	Hybrid	Resource Scheduling	CPU LB	Optimization	Not Specified	Low VM downtime, execution time	Low scalability and throughput
[50]	BFO+ Lamarack Evolutionary	Hybrid	Resource Scheduling	CPU LB	Optimization	Not Specified	Less transfer time	Low resource utilization
[51]	PSO	Dynamic	Task Scheduling	Task LB	Optimization	Not Specified	Low energy consumption	Low scalability
							High resource utilization	Low fault tolerance
								Small degree of balance
								Less makespan
								High SLV
[52]	GA	Dynamic	VM Scheduling	VM LB	Metaheuristic	Not Specified	Less response time	Low throughput
							Less makespan	Low scalability
							Less task rejection ratio	Small degree of balance
							Less task rejection ratio	Small resource utilization
[53]	GA	Dynamic	Task Scheduling	Task/VM LB	Optimization	G = O {n1 + (c × k) + (n2 + 1) (m + m + m)}	High degree of balance	Low scalability
							Less makespan	low energy efficiency
							Less execution time	low fault tolerance
							Less task rejection ratio	low fault tolerance
[54]	ACO and PSO	Dynamic	VM Scheduling	VM LB	Metaheuristic	O(n²MAI)	low response time	low throughput
							low execution time	low degree of balance
								high SLV
								low resource utilization
[55]	GA and GEL	Hybrid	Task Scheduling	VM LB	Optimization	Not Specified	high scalability	low degree of balance
							high fault tolerance	high SLV
							low overhead	low resource utilization
							low migration time and power consumption	high TRR
[56]	Honey Bee Algorithm	Dynamic	Task Scheduling	Task LB	Optimization	Not Specified	low response time	low throughput and scalability
							low execution time	low degree of balance
							low execution cost	low resource utilization
[57]	Non Classical	Dynamic	Resource Scheduling	Resource LB	Heuristic	Not Specified	High throughput	Low SLV
							High scalability	Low resource utilization
							Low response time	High task rejection ratio
							Low execution time	Low degree of balance
							Low execution time	High migration time
[58]	Non Classical	Dynamic	VM Scheduling	VM LB	Optimization	Not Specified	Low migration time	Low throughput
							High degree of balance	Low makespan
							Low response time	High SLV
								Low resource utilization
								Low scalability
[59]	BAT Algorithm	Dynamic	Resource/Task Scheduling	Resource/ Task LB	Optimization	Not Specified	Less execution time	High makespan
							Low execution cost	Low throughput
								Energy inefficient
								Low resource utilization
[60]	Non Classical	Dynamic	VM Scheduling	VM LB	Optimization	Not Specified	Less response time	Low scalability
							Low execution cost	High SLV
							Low execution cost	Low degree of balance
[61]	Simulated Annealing	Dynamic	Task Scheduling	Task LB	Optimization	Not Specified	High throughput	Low fault tolerance
							High scalability	Energy inefficient
							Low overhead	High SLV
							Less makespan
							High resource utilization
[62]	Round Robin	Static	VM Scheduling	VM LB	Heuristic	Not Specified	High Fault tolerance	Less scalability
							Small overhead	High SLV
							Less migration time	Low task rejection ratio.
							Good resource utilization	Low task rejection ratio.
[63]	Round Robin	Dynamic	Resource Scheduling	Resource LB	Heuristic	Not Specified	High Fault tolerance	Less scalability
							Less migration time	High SLV
							Good resource utilization	Low task rejection ratio
[64]	Non Classical	Static	Resource Scheduling	Resource LB	Optimization	Not Specified	Less Response time	Low throughput and scalability
							Low execution cost	Low resource utilization
							Low execution cost	Low degree of balance
[65]	Active Monitoring	Dynamic	VM Scheduling	VM LB	Heuristic	Not Specified	Less response time	Low throughput
							Less execution time	Low scalability
							Less execution cost	Low degree of balance
							Less execution cost	Low resource utilization
[66]	Active Monitoring	Dynamic	VM/Task Scheduling	VM/Task LB	Heuristic	Not Specified	High scalability	Low throughput
							Less response time	Low fault tolerance
							High resource utilization	High makespan
							High resource utilization	High SLV
[67]	Active Monitoring	Dynamic	Resource Scheduling	Resource LB	Heuristic	Not Specified	Low overhead	Low throughput
							Less makespan	Power inefficient
							High resource utilization	High SLV
[68]	Joint use of min-min and max-min	Static	Task Scheduling	Task LB	Optimization	Not Specified	High degree of balance	Low scalability
							Low makespan	Low fault tolerance
							Low execution time	High SLV
							High resource utilization	High SLV
[69]	Min-min	Static	Task/Resource Scheduling	Task/Resource LB	Optimization	Not Specified	Low makespan	Low throughput and scalability
							Low response time	High SLV and task rejection ratio
							High resource utilization	Power inefficient
[70]	Max-Min	Static	Task Scheduling	Task LB	Optimization	O (mn)²	High throughput and scalability	Low resource utilization
							Low fault tolerance	Low degree of balance
							Low overhead	High makespan
[71]	Round Robin	Dynamic	Task Scheduling	Task LB	Heuristic	Not Specified	Low makespan	Low fault tolerance
							Low power consumption	Low degree of balance
							Low SLV	Low resource utilization

Back to article page