The End-to-End Diagnostics Discovery (EDDY) project supports the integrated analysis and diagnosis of distributed, layered, and interdependent components and systems. EDDY provides a framework addressing the problem of end-to-end performance. It does this by defining a way to effectively bring together information about network activities and anomalies that allows integrated analysis, research, and auditing. Using EDDY-based tools, a system or network manager can more easily discover and diagnose problems as they occur, allowing independent processes to assist in their prediction, management, maintenance, and circumvention. An EDDY-enabled diagnostic infrastructure provides a significant new opportunity to address network performance problems.
A test version of the architecture and toolkit was released in 2006. The 0.5 Beta version of the EDDY Toolkit is intended to spur experimentation and innovation, and to encourage on-going broad participation in this effort. In conjunction with this first release, the EDDY Team fostered coordinated experimentation and leveraged learning in diagnostic infrastructures and analytics. Development of the EDDY Toolkit was supported with funding from Carnegie Mellon University, with additional support from Internet2 and the NSF Middleware Initiative (Cooperative Agreement No. OCI-0330626).