Understanding Internet Traffic Streams: Dragonflies and Tortoises
نویسنده
چکیده
We present the concept of network traffic streams, and the ways they aggregate into flows through Internet links. We describe a method of measuring the size and lifetime of Internet streams, and use this method to characterise traffic distributions at two different sites. We find that although most streams (about 45% of them) are dragonflies, lasting less than 2 seconds, a significant number of streams have lifetimes of hours to days, and can carry a high proportion (50% to 60%) of the total bytes on a given link. We define tortoises as streams that last longer than 15 minutes. We point out that streams can be classified not only by lifetime (dragonflies and tortoises) but also by size (mice and elephants), and note that stream size and lifetime are independent dimensions. We submit that Service Providers (ISPs) need to be aware of the distribution of Internet stream sizes, and the impact of the difference in behaviour between short and long streams. In particular any forwarding cache mechanisms in Internet routers must be able to cope with a high volume of short streams. In addition ISPs should realise that Long-Running (LR) streams can contribute a significant fraction of their packet and byte volumes – something they may not have allowed for when using traditional ‘flat rate user bandwidth consumption’ approaches to provisioning and engineering.
منابع مشابه
A multiscaling traffic model for UDP streams
Although most of the traffic carried over the Internet uses the Transmission Control Protocol (TCP) as the transport layer protocol, it is of paramount importance to develop models for streams that use the User Datagram protocol (UDP), since these streams are inelastic and, consequently, they can jeopardize the acquisition of bandwidth by TCP streams. This paper introduces a traffic model for U...
متن کاملCharacterizing Internet load as a non-regular multiplex of TCP streams
A commonly accepted traffic model for a large population of Internet users consists of a multiplex of Poisson-arriving heavy-tailed streams with the same constant rate (M/G/ ). We show that even though such regular model provides an accurate description of long-range dependence, the marginal distribution variance is underestimated, resulting in erroneous calculation of overflow probability in n...
متن کاملPRODUCED ON ACID - FREE PAPER MEASURING , UNDERSTANDING AND MODELLING INTERNET TRAFFIC Nicolas Hohn
This thesis concerns measuring, understanding and modelling Internet traffic. We first study the origins of the statistical properties of Internet traffic, in particular its scaling behaviour, and propose a constructive model of packet traffic with physically motivated parameters. We base our analysis on a large amount of empirical data measured on different networks, and use a so called semi-e...
متن کاملMeasuring P2P IPTV Systems
P2P IPTV systems start to be very popular on the Internet. Measuring the impact they have on the network and understanding their behavior is an important field. Available applications are based on proprietary solutions, thus traffic analysis is the only feasible way to identify the mechanisms used to get video streams. In this context, during the 2006 FIFA World Cup, we performed an extensive m...
متن کاملSome Observations of Internet Stream Lifetimes
We present measurements of stream lifetimes for Internet traffic on a backbone link in California and a university link in Auckland. We investigate the consequences of sampling techniques such as ignoring streams with six or fewer packets, since they usually account for less than 10% of the total bytes. We find that we often observe large bursts of small ‘attack’ streams, which will diminish th...
متن کامل