ABSTRACT
With the growth of high performance networking, a single host may have simultaneous connections that vary in bandwidth by as many as six orders of magnitude. We identify requirements for an automatically-tuning TCP to achieve maximum throughput across all connections simultaneously within the resource limits of the sender. Our auto-tuning TCP implementation makes use of several existing technologies and adds dynamically adjusting socket buffers to achieve maximum transfer rates on each connection without manual configuration.Our implementation involved slight modifications to a BSD-based socket interface and TCP stack. With these modifications, we achieved drastic improvements in performance over large bandwidth delay paths compared to the default system configuration, and significant reductions in memory usage compared to hand-tuned connections, allowing servers to support at least twice as many simultaneous connections.
- Cra92.Nettest, 1992. Network performance analysis tool~ Cray Research Inco]]Google Scholar
- FF96.Kevin Fall and Sally Floyd~ Simulations-based comparisons of tahoe~ reno and SACK TCPo Computer Communications Review, 26(3)9 July 1996~]] Google ScholarDigital Library
- Jac88.Van Jacobson~ Congestion avoidance and control~ Proceedings of A CM SIGCOMM '88, August 1988.]] Google ScholarDigital Library
- Mah96.Jamshid Mahdavi. Enabling high performance data transfers on hosts~ (notes for users and system administrators), November 1996~ Obtain via: http://www.psc.edu/networking/perf_tune.html.]]Google Scholar
- MBKQ96.Marshall Kirk McKusick, Keith Bostic, Michael J. Karels, and John S. Quarterman. The Design and Implementation of the ,~.~ BSD Operating System. Addison-Wesley, Reading MA, 1996.]] Google ScholarDigital Library
- Min97.March 1997. Private conversation between Greg Minshall and the authors.]]Google Scholar
- MM96.Matthew Mathis and Jamshid Mahdavi. Forward Acknowledgment: Refining TCP congestion control. Proceedings of A CM SIGCOMM '96, August 1996.]] Google ScholarDigital Library
- MSMO97.Matthew Mathis, Jeffrey Semke, Jamshid Mahdavi, and Teunis Ott. The macroscopic behavior of the TCP Congestion Avoidance algorithm. Computer Communications Review, 27(3), July 1997o]] Google ScholarDigital Library
- MSZ96.Qingming Ma, Peter Steenkiste, and Hui Zhango Routing high-bandwidth traffic in max-min fair share networks. Proceedings of ACM SIG- COMM '96, August 1996.]] Google ScholarDigital Library
- MTW98.Gregory J~ Miller, Kevin Thompson, and Rick Wilder~ Performance measurement on the vBNS~ In Interop'98 Engineering Conference, 1998.]]Google Scholar
- Net96.NetBSD 1.2 operating system, 1996. Based upon 4.4BSD Lite, it is the result of a collective volunteer effort. See http://www.netbsd.org~]]Google Scholar
- RFC793.J. Postel~ Transmission control protocol, Request for Comments 793, September 1981.]]Google Scholar
- RFC1191.Jeffrey Mogul and Steve Deering. Path MTU discovery, Request for Comments 1191, October 1991.]] Google ScholarDigital Library
- RFC1323.Van Jacobson, Robert Braden, and Dave Borman. TCP extensions for high performance, Request for Comments 1323, May 1992.]] Google ScholarDigital Library
- RFC2001.W. Richard Stevens. TCP slow start, congestion avoidance, fast retransmit, and fast recovery algorithms, Request for Comments 2001, March 1996.]] Google ScholarDigital Library
- RFC2018.Matthew Mathis, J amshid Mahdavi, Sally Floyd, and Allyn Romanow. TCP Selective Acknowledgement options, Request for Comments 2018, October 1996.]] Google ScholarDigital Library
- SAC98.Experimental TCP selective acknowledgment implementations, 1998. Obtain via: http://www.psc.edu/networking/tcp.html.]]Google Scholar
- Ste94.W. Richard Stevens. TCP//IP Illustrated, volume 1. Addison-Wesley, Reading MA, 1994.]]Google Scholar
- VS94.Curtis Villamizar and Cheng Song. High performance TCP in the ANSNET. ACM SIGCOMM Computer Communication Review, 24(5), October 1994.]] Google ScholarDigital Library
- Wel96.Von Welch~ A user's guide to TCP windows, 1996. Obtain via: http://www.ncsa.uiuc.edu/People/vwelch/net_perf/tcp-windows.html.]]Google Scholar
- WS95.Gary R. Wright and W. Richard Stevens. TCP//IP Illustrated, volume 2. Addison-Wesley, Reading MA, 1995.]]Google Scholar
Index Terms
- Automatic TCP buffer tuning
Recommendations
Automatic TCP buffer tuning
With the growth of high performance networking, a single host may have simultaneous connections that vary in bandwidth by as many as six orders of magnitude. We identify requirements for an automatically-tuning TCP to achieve maximum throughput across ...
Reno Friendly TCP Westwood based on Router Buffer Estimation
ICAS-ICNS '05: Proceedings of the Joint International Conference on Autonomic and Autonomous Systems and International Conference on Networking and ServicesTCP Reno versions are widely used in current network, however it has been actualized that their throughput deteriorates in high-speed network and wireless environments. To overcome these problems of TCP Reno versions, a number of protocols have been ...
TCP CERL: congestion control enhancement over wireless networks
In this paper, we propose and verify a modified version of TCP Reno that we call TCP Congestion Control Enhancement for Random Loss (CERL). We compare the performance of TCP CERL, using simulations conducted in ns-2, to the following other TCP variants: ...
Comments