ANALYSIS AND DISTRIBUTION OF SOCIAL NETWORK LOAD BY USING CLOUD SIM

Objectives: Is to analyze the necessity of a datacenter for a social network network to handle multiple user requests across multiple regions. Methods: The widely used social network, Facebook's usage statistics are collected for the year May 2019 in terms of the number of users actively using the social network, across the regions, and the data is further analyzed by distributing it over the network with a single datacenter and multiple datacentres. Multiple requests from the users if handled by a single datacenter is compared with handling the same number of requests using multiple datacenters. A cloud sim simulator is used to simulate a network to handle user requests. In the first scenario, the initial configuration of a datacenter in terms of VM Image Size, VM Memory, Bandwidth, Data Centre – Architecture, Data Centre – OS, Data Centre – Number of Machines, Data Centre – Number of processors per machine, Data Centre – VM Policy and other parameters is made and the result of handling multiple requests by the single datacenter is studied by recording the overall response time, datacenter's processing time, datacenter's service time and overall cost of the datacenter. In the 2nd scenario with the same configuration as of the first scenario for the datacenter is considered, but the multiple requests are handled by multiple datacenter and the results are recorded in terms of overall response time, datacenter's processing time, datacenter's service time and overall cost of a datacenter. Then a comparative analysis of the recorded results for a single and multiple datacenters is made. Findings: The performance of the multiple datacentres is better than the single datacenter to handle the multiple user requests from multiple regions in terms of response time and service time, but if multiple datacenters are used then the overall datacenter cost will be more compared to a single datacenter. Novelty/Application: Single/multiple datacenter analysis will help the social network networks to the user requests in an efficient way