Cluster Comput DOI 10.1007/s10586-017-1056-4 Dragonfly optimization and constraint measure-based load balancing in cloud computing Vijayakumar Polepally 1 · K. Shahu Chatrapati 2 Received: 12 April 2017 / Revised: 20 June 2017 / Accepted: 14 July 2017 © Springer Science+Business Media, LLC 2017 Abstract Load balancing is the significant task in the cloud computing because the cloud servers need to store avast amount of information which increases the load on the servers. The objective of the load balancing technique is that it maintains a trade-off on servers by distributing equal load with less power. Accordingly, this paper presents the load bal- ancing technique based on the constraint measure. Initially, the capacity and load of each virtual machine are calculated. If the load of the virtual machine is greater than the balanced threshold value then,the load balancing algorithm is used for allocating the tasks. The load balancing algorithm calculates the deciding factor of each virtual machine and checks the load of the virtual machine. Then, it calculates the selection factor of each task. Then, the task which has better selection factor is allocated to the virtual machine. The performance of the proposed load balancing method is evaluated with the existing load balancing methods, such as HBB-LB, DLB, and HDLB for the evaluation metrics load and capacity. The experimental results show that the proposed method migrate only three tasks while the existing method HDLB migrates seven tasks. Keywords Cloud computing · Load balancing · Constraint measure · Load · Capacity B Vijayakumar Polepally vijayakumarpolepally@gmail.com 1 Department of Computer Science and Engineering, Kakatiya Institute of Technology and Science, Warangal, Telangana, India 2 Department of Computer Science and Engineering, JNTUH College of Engineering, Manthani, Peddapalli, Telangana, India 1 Introduction Due to the advances in the Internet, the cloud computing has become an interesting research topic in recent years. The services provided by the cloud are accessed by the users, and the users did not need to perform any operations [1]. Cloud computing is defined as the process of providing services to the users, and it is mostly used in business and scientific domain. The users of the cloud access the services at anytime and anywhere in the world by pay-as-you-go style [2]. One of the vital tasks in the cloud computing is the load balancing [3–12] because the cloud servers need to store avast amount of information which increases the load on the servers. The objective of load balancing technique is that it maintains a trade-off on servers by distributing equal load with less power [13]. The characteristics of the cloud are different from the char- acteristics of other environments. The following are some of the characteristics of the cloud environments: The resources in the cloud are heterogeneous, and they are distributed in large areas. The operating environment and the instruments used by these resources are distinct from each other. The resources in the cloud are associated with each other through a dynamic communication layer and offer several services and accounting policies to the users [14]. The other charac- teristics [15] include (a) On-demand self-service: computing capabilities are provided by the consumer like network stor- age which did not require theinteractionof ahuman with the provider of the network service. (b) Broad network access: The services are presented on the broad network, and they are accessed by the several mechanisms. (c) Resource pool- ing: A multi-tenant model is used to pool the resources of the providers to the consumers and the resources, such as storage, memory, and so on are allocated to the consumers based on their needs. (d) Rapid elasticity: The services are 123