CASE FOR TCP VEGAS IN HIGH-PERFORMANCE COMPUTATIONAL GRIDS
- Eric H.
- Wu-Chun
Computational grids such as the Information Power Grid, Particle Physics Data Grid, and Earth System Grid depend on TCP to provide reliable communication between nodes across a wide-area network (WAN). Of the available TCP implementations, TCP Reno and its variants are the most widely deployed; however, Reno's performance in computational grids is mediocre at best. Due to conflicting results in the evaluation of TCP implementations we present a detailed simulation study that unifies the conflicting results and demonstrates the limitations of earlier work. We focus on the two most debated versions of TCP - Reno and Vegas. Using real trafJic distributions, we show that Vegas performs well over modern high-performance links and better than Reno with the proper selection of the Vegas parameters a and p. Our results exhibit ways to signijicantly enhance the performance of distributed computational grids that rely on TCP.
- Research Organization:
- Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
- Sponsoring Organization:
- USDOE
- OSTI ID:
- 975581
- Report Number(s):
- LA-UR-01-3420; TRN: US201018%%769
- Resource Relation:
- Conference: "Submitted to: IEEE International Symposium on High-Performance Distributed Computing."
- Country of Publication:
- United States
- Language:
- English
Similar Records
THE FAILURE OF TCP IN HIGH-PERFORMANCE COMPUTATIONAL GRIDS
Re-architecting flow-control adaptation for grid environments