Social networking sites have become an integral part in the modern world today. This thesis focuses on comparing and analyzing in detail the technical architectures of the two topmost leading networking sites--Facebook and LinkedIn. Detailed research on the communication services and network update services is presented. Scaling techniques: Functional partitioning, class partitioning along with data archiving and load balancing schemes have been explored in depth, and improvements have been proposed. The technical workflow between MySQL, Memcache, Alternative PHPcache, custom partitioning data, graph based data access libraries, leaf servers, aggregators, chat loggers, channel clusters, Network resolvers, Network collectors, Network digesters and supporting programming languages have been explored and documented. An extensive support to the open source platform with the customized developed frameworks has been well studied. Business aspects, strategies keeping the target audience in context, have been researched and then underlying opportunities in e-commerce world have been explored. An official survey was conducted with the approval from the Institutional Review Board among the target audience to learn about the users issues: following this appropriate solutions have been proposed. This thesis will serve as a single point of data access for the aspiring networking site developers to learn about technical architectures, corresponding supporting customized services, data models and all other required technical details of the two existing most popular networking sites.
