Load Balancing Algorithm over a Distributed Cloud Network

Due to rapidly expanding network bandwidth and hardware technologies, the Internet is developing at a rapid rate. Cloud computing, a relatively new concept, is a technique that achieves high reliability using low-power hosts. Cloud computing, an Internetbased development is dynamically scalable and often provides virtualized resources as a service over the Internet has become a significant development recently. Cloud computing essentially consists of a class of systems and applications that employ distributed resources to perform a function in a decentralized manner. The methodology of cloud computing is to utilize service nodes, often classified as computing resources on a network, to execute large and complicated tasks that require largescale computation. Proper selection of nodes for execution of a task is thus an important factor to exploit effective use of resources. The selection must be done according to the properties of the task. This study focuses on a two level task distribution system over a three tier cloud architecture. This paper studies the use of a hybrid task scheduling algorithm which combines two commonly used scheduling methods, the MM (Min-Min) and OLB (Opportunistic Load Balancing) to create our hybrid Balanced Load Min-Min algorithm (BLMM) algorithm. This leads to more efficient execution and maintains load balancing of the system nodes. KeywordsCloud Computing; Scheduling; Load Balancing; Hybrid Algorithm, BLMM