GBLOT: A Generator of BLOgosphere Traffic

Web 2.0 technologies have enabled the participation and collaboration of web users through emerging applications. Blogging is an example of Web 2.0 application for sharing informa-tion. These new applications induce workloads in the web with new features, whose study is important for introducing improvements at the architecture, operating system and middleware levels to allow a better support of them. The knowledge of workload characteristics is also essential for the development of workload generators that support capacity planning and traffic engineering activities. In this article, we describe GBLOT, a Generator of synthetic BLOgosphere Traffic based on analytical models, which can be used to evaluate the short-term behavior of blog hosting sites. The GBLOT's user session model is based on VBMGs (Visitor Behavior Model Graphs). The access patterns and transfer sizes of user sessions follow descriptive statistical models, based on characteristics of real traces. To validate GBLOT and demonstrate its usefulness, we generate traffic against artificial blogs managed by Movable Type publishing system. Our results show that the management of comments in Movable Type publishing system consumes an important amount of CPU time (above 70%).