Bufferbloat dataset

'’Bufferbloat’’ is the growth in buffer size that has led Internet delays to occasionally exceed the light propagation delay from the Earth to the Moon. Manufacturers have built in large buffers to prevent losses on Wi-Fi, cable and ADSL links. But the combination of some links’ limited bandwidth with TCP’s tendency to saturate that bandwidth results in excessive queuing delays. In response, new congestion control protocols such as BitTorrent’s uTP/LEDBAT aim at explicitly limiting the delay that they add at the bottleneck link.

In this page, we make our results, code and dataset available to the scientific community.

Experimental validation

We propose [PAM-13], demonstrate [PAM-12] and validate [GLOBECOM-13] a methodology for bufferbloat inference. The technique infeers the upstream queuing delay experienced by remote hosts: both those using LEDBAT, through LEDBAT’s native one-way delay measurements, and those using TCP, through the Time-stamp Option. The methodology has been published at [PAM-13], demonstrated at [P2P-12] and implemented in Tstat

We extensively validate the methodology in a testbed. To let the community repeat our results, we share the raw traces and logs that are used to obtain Fig. 2(a) and (b) shown above and described in this extended technical report. File names in the dataset are labeled accordingly.

• packet level traces
• kernel and application level ground truth logs

Internet experiments

To assess the extent of bufferbloat in the wild, we have performed an Internet measurement campaigns, whose data is available below. Specifically, we apply the technique on both active BitTorrent [TMA-13a] and large-scale scanning experiments [PAM-14b] to obtain measuremnts of Internet bufferbloat.

Our measurement campaign (see an excerpt in the figure below) let us conclude that:

• LEDBAT delay-based congestion control is effective in keeping the queuing delay low for the bulk of the peers,
• yet about 1% of peers often experience queuing delay in excess of 1 s, and
• not only the network access type, but also the BitTorrent client and the operating system concurr in determining the bufferbloat magnitudo.

We make the dataset used in [TMA-13] paper available to the community. Dataset consists of several BitTorrent packet-level traces, concerning 12 torrents, captured from 3 vantage points for a total of 88 experiments in which we contacted over 25K external peers.

• Internet experiments (20GB compressed size, 88 experiments, raw pcap traces captured)
• the four peers sample used in Fig.2 of [TMA-13]

References

• [P2P-12] , "Inferring the buffering delay of remote BitTorrent peers under LEDBAT vs TCP" IEEE P2P’XII sep. 2012, Conference Abstract Bibtex

  @inproceedings{DR:P2P-12,
    title = {Inferring the buffering delay of remote {BitTorrent} peers under {LEDBAT} vs {TCP}},
    author = {},
    year = {2012},
    booktitle = {IEEE P2P'XII},
    annote = {category=article state=published project=misc},
    note = {keyword=ledbat},
    month = sep,
    pages = {77-78},
    url = {http://www.telecom-paristech.fr/~drossi/paper/rossi12p2p.pdf},
    howpublished = {http://www.telecom-paristech.fr/~drossi/paper/rossi12p2p.pdf},
    www_txt_url = {http://www.telecom-paristech.fr/~drossi/paper/rossi12p2p.pdf}
  }
  

  Nowadays, due to excessive queuing, Internet delays grow sometimes as large as the propagation delay from moon to earth – for which the bufferbloat term was recently coined. Some points to active queue management (AQM) as its solution, others propose end-to-end congestion control techniques – like BitTor- rent that recently replaced TCP with the LEDBAT transport protocol. In this demo, we implement a methodology to monitor the upstream queuing delay experienced by remote hosts, both those using LEDBAT, through LEDBAT’s native one-way delay mea- surements, and those using TCP, through the timestamp option. By actively taking part into torrent downloads as leechers, our software is able to infer (and visualize) the amount of access delay suffered by the remote peers.
• [PAM-13] Chirichella, C. and Rossi, D. and Testa, C. and Friedman, T. and Pescape, A., "Remotely Gauging Upstream Bufferbloat Delays" Passive and Active Measurement (PAM), Extended Abstract mar. 2013, Conference Abstract Bibtex

  @inproceedings{DR:PAM-13,
    author = {Chirichella, C. and Rossi, D. and Testa, C. and Friedman, T. and Pescape, A.},
    title = {Remotely Gauging Upstream Bufferbloat Delays},
    booktitle = {Passive and Active Measurement (PAM), Extended Abstract},
    year = {2013},
    address = {Hong Kong, China},
    month = mar,
    annote = {category=article state=published project=mplane},
    url = {http://www.telecom-paristech.fr/~drossi/paper/rossi13pam.pdf},
    howpublished = {http://www.telecom-paristech.fr/~drossi/paper/rossi13pam.pdf},
    www_txt_url = {http://www.telecom-paristech.fr/~drossi/paper/rossi13pam.pdf},
    note = {keyword=ledbat,bufferbloat,mplane}
  }
  

  Bufferbloat is the growth in buffer size that has led Internet delays to occasionally exceed the light propagation delay from the Earth to the Moon. Manufacturers have built in large buffers to prevent losses on Wi-Fi, cable and ADSL links. But the combination of some links’ limited bandwidth with TCP’s tendency to saturate that bandwidth results in excessive queuing delays. In response, new congestion control protocols such as BitTorrent’s uTP/LEDBAT aim at explicitly limiting the delay that they add over the bottleneck link. This work proposes and validate a methodology to monitor the upstream queuing delay experienced by remote hosts, both those using LEDBAT, through LEDBAT’s native one-way delay measurements, and those using TCP (via the Timestamp Option).
• [TMA-13a] Chirichella, C. and Rossi, D., "To the Moon and back: are Internet bufferbloat delays really that large" IEEE INFOCOM Workshop on Traffic Measurement and Analysis (TMA’13) apr. 2013, Conference Abstract Bibtex

  @inproceedings{DR:TMA-13a,
    author = {Chirichella, C. and Rossi, D.},
    title = {To the Moon and back: are Internet bufferbloat delays really that large},
    booktitle = {IEEE INFOCOM Workshop on Traffic Measurement and Analysis (TMA'13)},
    year = {2013},
    address = {Turin, Italy},
    month = apr,
    annote = {category=article state=published project=mplane},
    url = {http://www.telecom-paristech.fr/~drossi/paper/rossi13tma-a.pdf},
    howpublished = {http://www.telecom-paristech.fr/~drossi/paper/rossi13tma-a.pdf},
    www_txt_url = {http://www.telecom-paristech.fr/~drossi/paper/rossi13tma-a.pdf},
    note = {keyword=ledbat,bufferbloat,mplane}
  }
  

  Recently, the \u201cbufferbloat\u201d term has been coined to describe very large queuing delays (up to several seconds) experienced by Internet users. This problem has pushed protocol designer to deploy alternative (delay-based) models to the standard (lossbased) TCP best effort congestion control. In this work, we exploit timestamp information carried in the LEDBAT header, a protocol proposed by BitTorrent as replacement for TCP data transfer, to infer the queuing delay suffered by remote hosts. We conduct a thorough measurement campaign, that let us conclude that (i) LEDBAT delay-based congestion control is effective in keeping the queuing delay low for the bulk of the peers, (ii) yet about 1% of peers often experience queuing delay in excess of 1s, and (iii) not only the network access type, but also the BitTorrent client and the operating system concurr in determining the bufferbloat magnitude.
• [GLOBECOM-13] Chirichella, C. and Rossi, D. and Testa, C. and Friedman, T. and Pescape, A., "Passive bufferbloat measurement exploiting transport layer information" IEEE GLOBECOM dec. 2013, Conference Abstract Bibtex

  @inproceedings{DR:GLOBECOM-13,
    title = {Passive bufferbloat measurement exploiting
    transport layer information},
    author = {Chirichella, C. and Rossi, D. and Testa, C. and Friedman, T. and Pescape, A.},
    year = {2013},
    booktitle = {IEEE GLOBECOM},
    annote = {category=article state=published project=mplane},
    month = dec,
    note = {keyword=traffic,ledbat},
    url = {http://www.telecom-paristech.fr/~drossi/paper/rossi13globecom.pdf},
    howpublished = {http://www.telecom-paristech.fr/~drossi/paper/rossi13globecom.pdf},
    www_txt_url = {http://www.telecom-paristech.fr/~drossi/paper/rossi13globecom.pdf}
  }
  

  "Bufferbloat" is the growth in buffer size that has led Internet delays to occasionally exceed the light propagation delay from the Earth to the Moon. Manufacturers have built in large buffers to prevent losses on Wi-Fi, cable and ADSL links. But the combination of some links limited bandwidth with TCP tendency to saturate that bandwidth results in excessive queuing delays. In response, new congestion control protocols such as BitTorrent uTP/LEDBAT aim at explicitly limiting the delay that they add at the bottleneck link. This work proposes a methodology to monitor the upstream queuing delay experienced by remote hosts, both those using LEDBAT, through LEDBAT native one-way delay measurements, and those using TCP, through the Timestamp Option. We report preliminary findings on bufferbloat-related queuing delays on an Internet measurement campaign involving a few thousand hosts.
• [PAM-14b] Casoria, P. and Rossi, D. and Auge, J. and Buob, Marc-Olivier and Friedman, T. and Pescape, A., " Distributed active measurement of Internet queuing delays " Passive and Active Measurement (PAM), Extended Abstract mar. 2014, Conference Abstract Bibtex

  @inproceedings{DR:PAM-14b,
    author = {Casoria, P. and Rossi, D. and Auge, J. and Buob, Marc-Olivier and Friedman, T. and Pescape, A.},
    title = { Distributed active measurement of Internet queuing delays },
    booktitle = {Passive and Active Measurement (PAM), Extended Abstract},
    year = {2014},
    address = {Los Angeles, USA},
    month = mar,
    annote = {category=article state=published project=mplane},
    url = {http://www.telecom-paristech.fr/~drossi/paper/rossi14pam-b.pdf},
    howpublished = {http://www.telecom-paristech.fr/~drossi/paper/rossi14pam-b.pdf},
    www_txt_url = {http://www.telecom-paristech.fr/~drossi/paper/rossi14pam-b.pdf},
    note = {keyword=bufferbloat,mplane,measurement}
  }
  

  Despite growing link capacities, over-dimensioned buffers are still causing, in the Internet of the second decade of the third millenium, hosts to suffer from severe queuing delays (or bufferbloat). While maximum bufferbloat possibly exceeds few seconds, it is far less clear how often this maximum is hit in practice. This paper reports on our ongoing work to build a spatial and temporal map of Internet bufferbloat, describing a system based on distributed agents running on PlanetLab that aims at providing a quantitative answer to the above question.