Faculty Publications
Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736
Publications by NITK Faculty
Browse
2 results
Search Results
Item Load balancing strategy for optimal peak hour performance in cloud datacenters(Institute of Electrical and Electronics Engineers Inc., 2015) Kulkarni, A.K.; Annappa, B.Cloud computing is a growing computing model that is influencing every other entity in the global business industry. Efficient load balancing techniques plays a major role in cloud computing by allocating requests to computing resources efficiently to prevent under/over-allocation of Virtual Machines (VMs) and improve the response time to clients. It is observed that during peak hours when request frequency is high, active VM load balancer (packaged in cloudAnalyst) over-allocates initial VMs and under-allocates later ones creating load imbalance. In this paper we propose a novel VM load balancing algorithm that ensures uniform allocation of requests to virtual machines even during peak hours when frequency of requests received in data center is very high to ensure faster response times to users. The simulations results suggest that our algorithm allocates requests to VM uniformly even during peak traffic situations. © 2015 IEEE.Item Cost aware service broker algorithm for load balancing geo-distrubuted data centers in cloud(Institute of Electrical and Electronics Engineers Inc., 2017) Kulkarni, A.K.; Annappa, B.With increased cloud adoption globally, the cloud service providers are setting up their data centers in various geographical location to cater the needs of diverse range of users across the globe. The cost of managing data center includes not only hardware, software costs but also the electricity costs prevailing at that location. The cost of electricity varies from location to location and it is mainly because of the uneven production & availability of resources, infrastructure to generate electricity at that part of the globe. It is important for data center owners to reduce the data center management cost without affecting the agreed SLA of service to its users. The paper proposes an algorithm for routing service requests to geo-distributed datacenters considering the electricity cost and response times to optimize the cost of datacenter management. The experimental results using cloud analyst show that our approach reduces the cost of data center management without any increase in the response time for users. © 2017 IEEE.