Diagnosing a TCP/ATM Performance Problem: A Case Study.

R. van Melle, C. L. Williamson and T. Harrison. Proceedings of IEEE GLOBECOM'97, Phoenix, AZ, pp. 1825-1831, November 1997.

Abstract: In May 1996, we obtained an SGI Indy workstation with OC-3 (155 Mbps) ATM connectivity to an experimental wide-area ATM network testbed. Unfortunately, our initial experiments with this network were disappointing: an abysmal end-to-end throughput of 4.3 Mbps for large file transfers between SGI Indy workstations at the University of Saskatchewan and the University of Calgary.

This paper describes our efforts in analyzing and diagnosing this performance problem, using cell-level measurements from the operational ATM network. Surprisingly, the performance problems are not due to send and receive socket buffer size problems, heterogenous ATM equipment, or cell loss within the ATM network. Rather, the poor performance is caused by an overrun problem at the sending workstation for the ftp transfer. The ensuing timeouts to recover from lost TCP segments accounts for the low end-to-end throughput achieved.