SETTLE


Space bEnchmarking and TesTing moduLEs!

For others will treat you as you treat them. Whatever measure you use in judging others, it will be used to measure how you are judged. Matthew 7:2 (NLT)

A tuple space is a shared memory component of middleware that provides communication and coordination facilities to the services in a distributed system. Tuple spaces have been used to implement a wide variety of applications, including parallel UNIX utilities, parallel genetic algorithms (GAs), distributed regression testing frameworks, large-scale mobile agent systems, ``lifestream'' information management applications, and scientific computations that support both astrophysics and bioinformatics research. Even though tuple spaces have been used to implement a wide range of applications, there is a relative dearth of benchmarking frameworks that focus on the measurement of tuple space performance. This research project is developing an approach to tuple space performance evaluation that supports the creation of concurrent local clients and the use of aging to populate the space with tuples before the execution of a benchmark. This research also provides the results from experiments that characterize the performance of a specific type of tuple space and evaluate the efficiency and effectiveness of the proposed aging technique.

Related Papers:

Related Presentations:

Project Team:

  • Gregory M. Kapfhammer (Allegheny),
  • Ahmed Amer (Pittsburgh)
  • Panos K. Chrysanthis (Pittsburgh)
  • Daniel Fiedler (Allegheny College)
  • Kristen Walcott (Allegheny College)
  • Thomas Richardson (Allegheny College)

This research project is supported by research and travel grants from the Department of Computer Science and the Dean's Office at Allegheny College. Please refer to Research for a complete listing of my current research activities.


Links to this Page