The RIPE Atlas Hackathon 2015

Please note that the deadline for applications to participate in this hackathon was 1 January 2015.

Event Information

This is a chance for developers, designers, computer science students, and open data enthusiasts to access the large open datasets produced by RIPE Atlas, a global Internet measurement network that measures Internet connectivity. Hackathon participants will be challenged to use this open data to develop useful, creative and stunning visualisations for the benefit of the entire Internet community.

Format

Small teams will work with RIPE Atlas developers and can choose to either work on furthering existing ideas developed by the RIPE Atlas team and community, or entirely new visualisations. At the end of the three-day event, we aim to have completed tools or working prototypes for all-new data visualisations.

Friday will include an introduction to RIPE Atlas and the formation of teams. On Saturday and Sunday, participants will work together in teams to further develop their ideas, culminating in short presentations to the jury on Sunday afternoon and a party to announce the winning teams and celebrate everyone's participation. Food and drinks will be provided throughout the three-day event.

All source code developed during the hackathon will be publicly licensed and available on GitHub, and available for the entire community to use.

Learn more, and see some examples of existing RIPE Atlas data visualisations, in this RIPE Labs article.

Details

  • Date: 27-29 March 2015
  • Time: Friday 12:00-18:00, Saturday 9:00-18:00, Sunday 9:00-21:00 (including party)
  • Location: Amsterdam, Netherlands
Christian Kaufmann

Measurement, Analysis and Tools (MAT) Working Group Chair

Elizabeth Turner

Data Designer, iconomical

Dr. Melanie Rieback

Co-founder/CEO, Radically Open Security

Jason Livingood

Vice President of Internet Services, Comcast

Robert Kisteleki

Research and Development Manager, RIPE NCC

Vesna Manojlovic

Technologia Incognita Hackerspace and Senior Community Builder, RIPE NCC

The three teams that produce the best results will each receive rewards of 1,500 EUR (to be split among non-RIPE NCC team members), courtesy of Comcast. Winning teams will also be promoted at the RIPE 70 Meeting in Amsterdam in May 2015. And all participants will receive RIPE Atlas t-shirts!

Comcast will provide some travel funding for a number of participants, depending on their financial need, their background and experience, and their reasons for wanting to participate in the hackathon. The organisers and jury will make final decisions about who will be awarded travel funding and how much funding participants will receive.

For Participants

We look forward to hosting you for the first-ever RIPE Atlas hackathon in Amsterdam from 27-29 March!

Here is some more information about the kinds of challenges you'll be working on during this three-day event.

Goal

Visualising RIPE Atlas data in useful, new creative ways.

Challenges

  • Track 1: Create a working prototype of a stand-alone application using any technology you want (aka go crazy, surprise us!)
  • Track 2: Use technology that can be easily integrated into the RIPE Atlas system and maintained by RIPE Atlas developers (aka stuff we can take over) - see more about the requirements below
  • Track 3: Integrate RIPE Atlas data into existing software (aka expose RIPE Atlas data to the world)

Criteria

Required

  • Visually compelling view of the data
  • Application and usefulness for the community (network operators, researchers, analysts, general public)
  • Solving one or more use cases (see below)
  • Code is flexible and reusable
  • Visualisation adjustable to the regional level (continental, RIR service regions, countries, etc.)
  • Multiple datasets used

Use Cases

Based on our own ideas and requests we've received from members of the community, the following are a few "use cases" that could help get you started:

  • Comparing IPv6 and IPv4 reachability
  • Correlate BGP (RIS) and RIPE Atlas data
  • Visualise "net-quake": view of an event depending on the distance from the "epicentre" of the disturbance
  • Show how peering improves performance
  • Make something cool with streaming data
  • SSL Watch
  • Google Earth view
  • Integrating RIPE Atlas with IXP Manager

Your project must be licensed by an OSI-approved license, but here are some of our favourites:

Available Data and Resources

In addition to the resources below, hackathon participants will also have access to RIPE Atlas developers, who will be available throughout the three days.

RIPE NCC open datasets

RIPE Atlas APIs and Documentation

All RIPE Atlas documentation is available online. In particular, the following may be useful:

Challenge Track 2

Existing Code on GitHub

Requirements for integration into the RIPE Atlas system

  • Code must:
    • Be written in Python if you require server-side development
    • Be written in JavaScript (or an established JS framework) for client-side development.
  • Code must not:
    • Make use of proprietary technologies like Flash or Silverlight
    • Be precompiled
    • Be obfuscated

Examples of previous work

Helpful tools

More detailed ideas

These are just some ideas that may be of interest to you or could help get the ball rolling - but don't feel limited! There are endless ways to use RIPE Atlas data to create interesting visualisations, and we're looking forward to seeing your own ideas, too.

  1. Track general DNS recursive resolver performance over time, and be able to compare two or more continents/countries/ISPs. For example, it may be interesting to visualise what DNS query response times look like in Africa vs. South America, or between ISP #1 and ISP #2 in a given country.
  2. Visualisation of IPv6-enabled probes. This could track the percent of IPv6-enabled probes by network, country, and continent and enable comparative reporting (e.g. Korea is at 29% and Thailand is at 12%, while Sweden is at 87%).
  3. Comparative IPv4 vs. IPv6 measurements performance. Answer the question of whether IPv6 is really faster/better.
  4. Create a live-update table that shows the number of observed root servers in each content and country, or the number of instances of each root server over time.
  5. A traceroute visualisation with a variable timeframe that can show actual routing changes over time for any traceroute measurement.
  6. A measurement and visualisation to display the reachability and general performance of top CDNs, as observed from different regions of the world.
  7. A package for R (a commonly used statistical computing and analysis environment, especially popular in academia) that makes it easy to import and use RIPE Atlas data for statistical analysis. The package could include the following capabilities:
    • Documentation of the data (definitions of the variables in the dataset)
    • Summary of the size of the available data: how many rows (records) and columns (variables) there are
    • Ability to import all available RIPE Atlas data into R as a data frame
    • Ability to import a subset of the available data into R as a data frame:
      • Random sample of the data
      • Filtered by range of date and/or by geographical location, etc.