IWCA 2012 - International Workshop on Cloud Computing Applications (IWCA)

Cloud Computing is undoubtedly becoming an emerging platform for both the industrial and scientific community. The promise of scale, elasticity, and “pay-as-you-go” pricing models is attracting corporate IT staff, researchers, and individual software application developers. With the advent of “Platform-as-a-service” or PaaS, there is significant support for design, build, test, and deployment tools in the cloud. Organizations the world over are spending money and effort on experimenting with whether Cloud is suitable for new application development and if it is mature enough for migrating existing applications. However, Cloud is not just another new platform to write applications on. We believe that to harness the true potential of Cloud in terms of scalability and agility, one needs to design and implement the applications in new ways. For example, the traditional Apache server model of process-per-connection (PPC) is not scalable enough to handle the load bursts that cloud-based websites face today (Slashdot effect) and as a result, an event-driven server architecture is needed. Similarly, very fast content write and lookup requirements of Extreme Transaction Processing (as in Twitter, Facebook, and similar social networking sites) have given rise to new genre of Key-Value stores with faster performance and more efficient access. No SQL databases in the Cloud are optimized for varied special purpose requirements, such as search and range queries (compared to RDBMS databases which are optimized for serving structured queries only). Today’s cloud-scale data processing frameworks like Hadoop for map-reduce also bring newer dimensions to batch-processing frameworks. Furthermore, PaaS platforms introduce the paradigm of multi-tenancy at the application level with their own set of challenges by imposing resource quotas.