Grid and distributed computing environments are evolving rapidly and driving the development of system and network technologies. The design of applications has placed an increased emphasis upon adapting application behavior based on the performance of the network. In addition, network operators and network researchers are naturally interested in gathering and studying network performance information. This work presents an extensible framework for the storage and exchange of performance measurements. Leveraging existing storage and exchange mechanisms, the proposed framework is capable of handling a wide variety of measurements while delivering performance comparable to that of less flexible, ad-hoc solutions.
Jason Zurawski, D. Martin Swany, Dan Gunter