TCP-friendly marking for scalable best-effort services on the internet
Giampiero Lo Monaco, Azeem Feroz, Shivkumar Kalyanaraman, Yong Qing Xia
Abstract
Giampiero Lo Monaco, Azeem Feroz, Shivkumar Kalyanaraman, Yong Qing Xia
Abstract
This paper proposes the use of TCP-aware network-based packet marking, which in conjunction with differential packet dropping, is a powerful way to improve the performance of buffer management for best-effort services on the Internet. We extend the notion of "TCP-Friendly" packet marking, proposed recently by us [6], and apply it to improve the performance of traditional best-effort services. The TCP-aware use of deterministic packet marking at the network edge allows us to protect selected TCP packets from suffering loss. This significantly reduces the total number of timeouts and avoids the resulting service degradation. In particular, we protect TCP sessions with very small windows, disperse packet loss across a given window, and protect retransmitted packets from encountering loss. We show baseline results which illustrate that the performance gains are considerable (orders of magnitude) when compared to stateful packet dropping algorithms like FRED [15], and even when TCP SACK [18] implementations are employed. The scheme has been implemented in Linux 2.2.10, and all the results are based upon experimental data.
OpenAlex reports 12 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper proposes the use of TCP-aware network-based packet marking, which in conjunction with differential packet dropping, is a powerful way to improve the performance of buffer management for best-effort services on the Internet. We extend the notion of "TCP-Friendly" packet marking, proposed recently by us [6], and apply it to improve the performance of traditional best-effort services. The TCP-aware use of deterministic packet marking at the network edge allows us to protect selected TCP packets from suffering loss. This significantly reduces the total number of timeouts and avoids the resulting service degradation. In particular, we protect TCP sessions with very small windows, disperse packet loss across a given window, and protect retransmitted packets from encountering loss. We show baseline results which illustrate that the performance gains are considerable (orders of magnitude) when compared to stateful packet dropping algorithms like FRED [15], and even when TCP SACK [18] implementations are employed. The scheme has been implemented in Linux 2.2.10, and all the results are based upon experimental data.
Key concepts: Computer science, Computer network, Packet loss, Stateful firewall, Network packet, TCP global synchronization, TCP acceleration, Scalability