1、Amogh Dhamdhere,1,Internet Traffic Characterization,Amogh Dhamdhere,Amogh Dhamdhere,2,What is covered in this talk,Why characterize Internet traffic ? Measurement and analysis methodologies. Measurement studies. Variation of Internet traffic (time of day, day of week effects) Packet level characteri
2、stics (packet sizes). Flow level characteristics (Flow sizes, flow durations). File size distributions. Distribution by application. Distribution by protocol.,Amogh Dhamdhere,3,What is not covered,Everything that will be covered in future presentations ! Delay and loss measurements TCP related measu
3、rements (TCP flavors etc) Self similarity of Internet traffic Flow measurements Peer to peer traffic measurements,Amogh Dhamdhere,4,Goals of this research,Observe Internet traffic characteristics. Develop reasonable models to understand these characteristics. Failure of traditional mathematical mode
4、ling techniques (e.g. Queueing theory). Earlier models deal with issues which are non-critical from the practitioners point of view. Attempt to close the void between theory and practice.,Amogh Dhamdhere,5,Why Characterize Internet Traffic ?,Provisioning network resources (capacity, buffer, etc) How
5、 should the network be provisioned to satisfy certain constraints. Constraints may differ with the type of traffic. E.g. Buffer provisioning Current tools (eg SNMP) may not be sufficientAnalyzing network performance TCP performance Routing performance,Amogh Dhamdhere,6,Why Characterize Internet Traf
6、fic ?,Obtain characteristic workloads for use in simulations Typical packet sizes Typical flow durations Most commonly used TCP flavorsImportant for ISPs to formulate policy decisions (Service Level Agreements)Developing techniques to detect network anomalies e.g. Denial of Service attacks.Verify ru
7、le of thumb type design guidelines.,Amogh Dhamdhere,7,Measurement Methodologies,Objectives of a monitor: Collection of detailed traffic statistics from heterogeneous network links. Non-interference with the measured network (non-intrusiveness). Obtaining a global view of the monitored network from a
8、 reasonable number of monitoring points.Types of monitor: Active monitors Passive monitors,Amogh Dhamdhere,8,IPMON (Sprint),Passive monitor for the Sprint backbone network. Capable of monitoring links of capacities ranging from OC-3 to OC-48. Uses an optical splitter on the monitored link. Records p
9、acket traces including IP and TCP/UDP headers, timestamp. Trace sanitizer. Analysis component: Flow statistics (start and end time of flows, flow sizes) Protocol (TCP, UDP) and application (web, email, streaming) split of traffic.,Amogh Dhamdhere,9,IPMON,Amogh Dhamdhere,10,Other Projects,OC3MON (MCI
10、) - Passive monitor designed for OC3 links (155 Mbps). NetScope (AT&T) - A set of tools for traffic engineering in IP backbone networks. Network Analysis Infrastructure (NAI) - Performance of vBNS (very high speed Backbone Network Service) and Abilene networks. Some routers have built-in monitoring
11、capabilities. Netflow Cisco routers.Commercial tools Niksuns NetDetector and NikScouts ATM Probes.,Amogh Dhamdhere,11,Measurement Studies,Wide Area Internet Traffic Patterns and Characteristics Thompson, Miller, Wilder, MCI Telecommunications, 1997.One of the first studies of commercial backbone tra
12、ffic. Used the OC3MON traffic monitor described earlier, at two locations on MCIs commercial backbone. Characterize traffic on timescales of 24hrs and 7 days in terms of traffic volume, flow volume, flow duration, packet sizes, traffic composition (by protocol, application). Two links monitored. Dom
13、estic and International.,Amogh Dhamdhere,12,MCI Study Daily and weekly effects,Traffic volume shows a clear diurnal pattern, with traffic tripling from 06:00 through 12:00 noon EDT. Traffic decreases by about 25% during the weekend. The two directions of the monitored link are not symmetric.,Amogh D
14、hamdhere,13,MCI Study Asymmetry in packet sizes,Packet sizes are different in the two directions, and are roughly inversely proportional to each other.,Amogh Dhamdhere,14,MCI Study Packet size distributions,Packet size distributions are trimodal. 40-44 bytes - TCP ACKs, control segments etc. 552 or
15、576 bytes - Default MSS when MTU Discovery is not used is 512 or 536 bytes. 1500 bytes MTU for Ethernet.,Amogh Dhamdhere,15,MCI Study International Link Traffic,International link traffic shows similar time of day, day of week effects. Packet sizes in the two directions are asymmetric Larger packets
16、 in the U.S. to U.K. direction.,Amogh Dhamdhere,16,MCI Study Protocol and Application Mix,Protocol composition TCP dominates (95% of bytes, 90% packets, 75% flows) UDP second (5% bytes, 10% packets, 20% flows) ICMP most of the remaining. Application composition Web (75% bytes, 70% packets, 75% flows
17、) Other (may also be web-related) DNS (1% bytes, 3% packets, 18%) SMTP (5% bytes, 5% packets, 2% flows) FTP (5% bytes, 3% packets, 1% flows) NNTP (2% bytes, 1% packets, 1% flows) Telnet (1% bytes, 1% packets, 1% flows),Amogh Dhamdhere,17,Measurement Studies,Trends in Wide Area IP Traffic Patterns Mc
18、Reary, Claffy, CAIDA, 2000.Data collected by the NAI project from May 1999 through March 2000 at the NASA Ames Internet Exchange. Analysis of packet size distributions, protocol/application mix etc. Show increasing trends in traffic from new (at that time) applications e.g. streaming media, online g
19、ames, Peer to Peer (Napster). No change in the overall trend in the TCP/UDP traffic ratio as compared to the analyses at MCI and CAIDA in 1998.,Amogh Dhamdhere,18,CAIDA Study Packet Size Distributions,Packet size distributions show same trimodal trend as previous results.,Amogh Dhamdhere,19,CAIDA St
20、udy Protocol and Application Mix,Protocol mix TCP and UDP are still the most popular protocols, and in roughly the same proportions.Application mix (TCP) Web is still the most popular application New applications like peer to peer file sharing (Napster) now appear in the list. (Napster at 5th positi
21、on)Application mix (UDP) Streaming media (RealAudio) now comprises a substantial portion of total UDP traffic. Online games (Half Life, EverQuest, Unreal, Quake 3) also have substantial share.,Amogh Dhamdhere,20,CAIDA Study Long Term Trends,The protocol mix of the traffic (TCP and UDP) does not chan
22、ge significantly over time. Decline in the contribution of FTP to the overall traffic mix. Possibly due to shift from active to passive mode FTP, because of an increase in packet filtering firewalls. Alternate protocols for file transfer. Decline in the fraction of RealAudio traffic. RealAudio traff
23、ic has remained fairly constant, while other traffic has increased. Decline in the fraction of game traffic,Amogh Dhamdhere,21,CAIDA Study Long Term Trends,Significant increase in peer to peer traffic (Napster),Amogh Dhamdhere,22,CAIDA Study Short Term Trends,Email traffic increased significantly in
24、 November and early December, decreasing after December holidays.,Amogh Dhamdhere,23,CAIDA Study Short Term Trends,Online gaming shows day of week effects, with traffic nearly doubling over weekend periods.,Amogh Dhamdhere,24,Measurement Studies,Longitudinal study of Internet traffic from 1998-2001
25、Fomenkov, Keys, Moore, Claffy, CAIDA, 2001.Unique long term view of Internet traffic. Multiple observation sites (20) Four metrics of measured traffic Number of bytes. Number of packets. Number of flows. Number of source-destination pairs (port number and protocol fields ignored). This measures the
26、number of Internet hosts communicating via the monitored link.,Amogh Dhamdhere,25,Longitudinal Study,Bit and packet rates show diverse behavior Some sites show sustained growth, some are constant and some fluctuate between growth and reduction. No clear diurnal pattern in the measured traffic ! No c
27、onsistent long term growth Refutes the notion that Internet traffic ic universally and rapidly increasing.Usage patterns Traffic composition varies significantly from site to site. WWW traffic reached maximum between late 1999 and early 2000. Has been constant or decreased since. This could be due t
28、o the onset of noticeable amounts of P2P traffic.,Amogh Dhamdhere,26,Longitudinal Study Application Mix,Amogh Dhamdhere,27,Measurement Studies,Packet Level Traffic Measurements from the Sprint IP Backbone Fraleigh, Moon, Lyles, et al. Sprint Labs, 2003Most recent (2001-2002) study of traffic on a co
29、mmercial backbone link. Analyses the impact of new applications (distributed file sharing, streaming media) New results for end-to-end loss and delay performance of TCP connections. Measurements of network delays in the backbone and U.S. transcontinental links. Methodology Uses the IPMON architectur
30、e described earlier.,Amogh Dhamdhere,28,SPRINT Study Traffic Load,Traffic load in bytes SNMP is not able to capture the burstiness of the traffic at smaller timescales.Most backbone links are utilized under 50%. Less than 10% of the backbone links experience utilization higher than 50% in any 5 min
31、interval. Noticeable peaks in traffic load are observed due to DoS attacks.Traffic in a bidirectional link is asymmetric. Many applications are inherently asymmetric. Hot potato routing.,Amogh Dhamdhere,29,SPRINT Study,SNMP is not able to capture the burstiness of the traffic at smaller timescales.,
32、Amogh Dhamdhere,30,SPRINT Study Application Mix,Application mix varies from link to link. In most cases, web represents more than 40% of total traffic (As seen in previous studies) However, on some links, the web contributes less than 20%, while P2P accounts for 80%. Streaming applications are a sta
33、ble component of the traffic.,Amogh Dhamdhere,31,SPRINT Study - Flows,The number of flows and the traffic load are not necessarily correlated. i.e a large number of flows does not always mean a large traffic load.,Amogh Dhamdhere,32,Measurement Studies Flow level,Understanding Internet Traffic Strea
34、ms: Dragonflies and Tortoises Brownlee, Claffy CAIDA. Results of flow level measurements from two links: OC3 link (Auckland) and OC12 link (UCSD) Uses an extension of NeTraMet to monitor stream lifetimes. Previous classifications of flows were on basis of size (packets or bytes) Elephants (large tra
35、nsfers) Mice (short transfers) Propose alternate classification of TCP flows on basis of their lifetime. Tortoises (long lasting transfers) Dragonflies (short duration transfers) Here flows are defined as sets of packets traveling in either direction between a pair of end-points.,Amogh Dhamdhere,33,
36、Dragonflies and Tortoises,Percentages of streams and bytes. Long Running (LR) streams (15 mins) account for about 1% of the streams. Very Short streams (2 sec) account for 40 70 % of streams, showing a diurnal pattern of variation. At UCSD site, 50% of all bytes were in LR streams, while this fracti
37、on was 5% for Auckland. Most of these streams are non-web traffic.,Amogh Dhamdhere,34,Short Streams Streams lasting less than 15 mins,Lifetime distributions 45% of streams have lifetimes less than 2 sec. Distributions do not change rapidly over time.,Amogh Dhamdhere,35,Short Streams Streams lasting
38、less than 15 mins,Byte size distributions Short stream size distributions for UDP, non-web TCP and web TCP are considerably different. Distributions are stable over long periods of time,Amogh Dhamdhere,36,Tortoises Streams lasting more than 15 mins,Bit rates Longer duration LR streams are low-rate (
39、interactive) or high rate (multimedia) with approximately equal frequency. Medium duration LR streams tend to be high-rate. (file transfers) UDP streams run at constant bit rates, but these rates may change in response to the applications state (online games).,Amogh Dhamdhere,37,Tortoises Streams la
40、sting more than 15 mins,LR stream lifetimes LR stream lifetimes seem to follow a power law distribution.,Amogh Dhamdhere,38,Measurement Studies Flow level,Internet Stream Size Distributions Brownlee, Claffy, CAIDA 2002.Measurements of Per minute distributions of stream sizes in bytes for a period of
41、 one hour. Two different types of traffic considered: Web traffic, and non-web TCP traffic.Web streams 87% under 1kB, 8% between 1 and 10 kB, 4.8% between 10 and 100 kB.Non-web streams 89% under 1kB, 7% between 1 and 10 kB, 1.5% between 10 and 100 kB.,Amogh Dhamdhere,39,Internet Stream Size Distribu
42、tions,Amogh Dhamdhere,40,File Size Distributions,The Structural cause of file size distributions Downey, 2001. A new model for the operations that create new files. Files appear because of common operations. Copying. Translating and filtering. Editing. Using this, the distribution of file sizes can
43、be predicted to be lognormal. Start with a single file of size s*. Select a file size s at random from the current distribution. Create a new file with size fs and add to the distribution. (f is a factor chosen from some other distribution. Hence size of nth file is sn = s* f1 f2 f3fmlog(sn) = log(s
44、*) + log(f1) + .,Amogh Dhamdhere,41,File Size Distributions,File sizes on web servers Studies by Arlitt and Williamson claim file size match the Pareto model. This may not be true ! Some of the analyzed data sets better fit the lognormal model.Traces of downloaded files. Fits a hybrid model with log
45、normal distribution with a Pareto tail. Two mode lognormal model is also a good match. Summary The distribution of file sizes is NOT heavy tailed ! Implications on self-similarity of Internet traffic Most explanations assume that distribution of file sizes is long-tailed. Need to revise explanations
46、 of self-similarity.,Amogh Dhamdhere,42,Non-commercial networks,Some results from the abilene network during the duration of one week.Application mix Web traffic is much lower as compared to commercial backbone networks. Email traffic is higher. Measurement traffic amounts to 5% of all traffic !Prot
47、ocol mix TCP is still the most dominant (90% of bytes). UDP accounts for 5%. ICMP around 4%. Numbers similar to that on commercial backbone links.,Amogh Dhamdhere,43,Future Directions,Self-similarity The need to verify assumptions. Downey questioned the assumptions about file size distributions. Int
48、er-arrival time distributions. Transfer length distributions. Burst size distributions. Dependence of traffic characteristics on TCP algorithms.Measurement based forecasting of DoS attacks and flash crowds.Real time monitoring of critical parameters. Use this characterization to automatically make decisions. Provisioning. Routing etc.,Amogh Dhamdhere,44,Future Directions,Characterization of P2P traffic. Previous measurement studies on P2P systems focused on node behavior, topology etc. Need to better characterize the traffic generated by P2P applications.,Amogh Dhamdhere,45,Thank You !,
