Is Explicit Congestion Notification usable with UDP?

McQuistin, Stephen; Perkins, Colin

doi:10.1145/2815675.2815716

Cited by 10 publications

(7 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition to adding data points to this and our own previous work, we revise our estimate of pathdependent impairment in ECN, and therefore the proportion of connection attempts that need more than simple fallback to deal with in-network impairment. More recently, McQuistin et al [14] complemented this data also with measurements of ECN support for UDP traffic. Automating PATHspider measurements has allowed us to collect more data, and better isolate transient effects from actual path dependency.…”

Section: Related Workmentioning

confidence: 90%

Tracking transport-layer evolution with PATHspider

Trammell

Kühlewind

Vaere

et al. 2017

Proceedings of the Applied Networking Research Workshop

View full text Add to dashboard Cite

The ossification of the Internet protocol stack, due in large part to mangling of packets by middleboxes, has led to a relatively slow rate of change in today's Internet. We have developed the PATHspider active Internet measurement tool which performs onesided measurements of a variety of transport-layer features and extensions, to investigate these impairments to protocol evolution along an Internet path. Data collected with PATHspider can be used both to determine the degree of support for these features, as well as to detect connectivity issues caused by attempting to use them. The wider aim of this effort is to provide quantifiable input to protocol design and deployment choices that can be based on the level of impairment present in the Internet.This paper details PATHspider's design, and applies it to trace the evolution of deployment of two extensions to TCP, Explicit Congestion Notification (ECN) and the newer TCP Fast Open (TFO); as well as the degree of interference with the Differentiated Services Code Point (DSCP) carried in the IP header. Our ECN results, in particular, expand on a long-term study beginning in 2012, and show continued linear adoption of ECN. Automating PATHspider measurements has allowed us to collect far more data than in previous campaigns, allowing us to better distinguish ECN-linked connectivity failures from transient effects. Interestingly, we observe a correlation between ECN-linked connectivity failure in the core of the network with the presence of large-scale, heterogeneous Internet censorship infrastructure.

show abstract

Section: Related Workmentioning

confidence: 90%

Tracking transport-layer evolution with PATHspider

Trammell

Kühlewind

Vaere

et al. 2017

Proceedings of the Applied Networking Research Workshop

View full text Add to dashboard Cite

show abstract

“…UDP traffic constitutes a non-trivial share of Internet traffic in terms of telephony, media streaming, teleconferencing, and more, reinforced by the recent surge of people working remotely from home [51]. In addition, using UDP as the basis for designing new transport protocols is common [14]. In the second scenario (Figures 13b and 14b), to further investigate the performance inefficiencies in the presence of bursty UDP traffic, we expose the same foreground data transfer to more dynamic UDP background traffic patterns.…”

Section: Impact Of Cross Traffic On High-performance Toolsmentioning

confidence: 99%

“…We show that if there is a significant amount of UDP background traffic, especially bursty UDP traffic, GridFTP is negatively affected. The mix of TCP and UDP traffic on shared networks do change over time, but there are emerging application protocols that do use UDP (e.g., [14,15]) For example, our empirical results show that if the competing background traffic is UDP-based, then UDT (which is UDP-based itself) can actually have significantly higher throughput (Figure 13a, BG-SQ-UDP1 and BG-SQ-UDP2; Figure 13b all combinations except for No-BG).…”

mentioning

confidence: 99%

Machine-Learned Recognition of Network Traffic for Optimization through Protocol Selection

Anvari

2021

Computers

View full text Add to dashboard Cite

We introduce optimization through protocol selection (OPS) as a technique to improve bulk-data transfer on shared wide-area networks (WANs). Instead of just fine-tuning the parameters of a network protocol, our empirical results show that the selection of the protocol itself can result in up to four times higher throughput in some key cases. However, OPS for the foreground traffic (e.g., TCP CUBIC, TCP BBR, UDT) depends on knowledge about the network protocols used by the background traffic (i.e., other users). Therefore, we build and empirically evaluate several machine-learned (ML) classifiers, trained on local round-trip time (RTT) time-series data gathered using active probing, to recognize the mix of network protocols in the background with an accuracy of up to 0.96.

show abstract

“…Further, there have been a few recent studies that also focus on UDP, e.g. on ECN support [18], complementing our earlier ECN measurement, or our own measurement on UDP differential treatment [19] which was performed together with some of the authors of the tracebox tool [20]. Further, DSCP modification in the IP header was recently examined [21], however, only providing initial results from a small-scale measurement study.…”

Section: A Related Workmentioning

confidence: 99%

Tracing Internet Path Transparency

Kühlewind

Walter

Learmonth

et al. 2018

2018 Network Traffic Measurement and Analysis Conference (TMA)

View full text Add to dashboard Cite

Investigating Internet Path Transparency means measuring if a network path between two endhosts is impaired by in-network functions on the path. A path is considered transparent if it provides connectivity and the same performance independent of the protocol or protocol stack that is used for the transmission. Unfortunately this is not always the case. Simple firewalls that block e.g. UDP, are an example. Of course such in-network functions are often valuable, like firewalls. However, these middleboxes also, sometimes unintentionally, make assumptions about the traffic passing through them that restricts innovation in the Internet on the higher layers, e.g. the deployment of new UDP-based protocols such as QUIC, to stick with the previous example. PATHspider is an active measurement tool to test for Path Transparency. In this paper we present a new feature of PATHspider that integrates tracebox-based functionality and analysis to not only detect in-transparency but also further locate the origin of the impairment observed. As an example study we show updated and extended measurements on ECN support and connectivity. By using our enhanced ECN PATHspider plugin to test network support of the ECN IP codepoint and additional path tracing that is correlated with DSCP testing, we show that most in-network ECN IP codepoint zeroing is due to use of the deprecated definition of the IP ToS field for domain-internal service differentiation, while pure resetting of the ECN IP field is more likely an active inference in border networks.

show abstract

Is Explicit Congestion Notification usable with UDP?

Cited by 10 publications

References 11 publications

Tracking transport-layer evolution with PATHspider

Tracking transport-layer evolution with PATHspider

Machine-Learned Recognition of Network Traffic for Optimization through Protocol Selection

Tracing Internet Path Transparency

Contact Info

Product

Resources

About