Hello, We are excited to announce the release of our tool, D-Tunes, which helps application developers configure replication in geo-distributed datastores like Cassandra. D-Tunes is a research prototype that we have developed at the Internet System Laboratory (ISL) at Purdue University, and is part of our overall effort on architecting latency sensitive web applications for the cloud.
D-Tunes helps application developers configure replication parameters like number of replicas, location of replicas and quorum parameters in the best possible way to ensure that application constraints like low latency, stronger consistency and availability can be met. While the underlying techniques are general, the current release of D-Tunes targets Cassandra deployments on Amazon EC2 given the popularity of these platforms. Our tool could also be useful for estimating the response latency that can be expected if a single data-center deployment of Cassandra were migrated to a geographically distributed multi-region setup. The current version of D-Tunes can be used from the web-portal here: http://dtunes.ecn.purdue.edu/cgi-bin/home.py We would love to hear any feedback, questions or suggestions that you might have for us, which will be immensely helpful to us in guiding our research, and developing the next versions of D-Tunes. Please use the feedback form on the web-site, or email as at spuzh...@purdue.edu. D-Tunes Project Team Internet Systems Lab, Purdue University.
