About Measurement Lab

Measurement Lab (M-Lab) is an open, distributed server platform for researchers to deploy Internet measurement tools. The goal of M-Lab is to advance network research and empower the public with useful information about their broadband connections. By enhancing Internet transparency, M-Lab helps sustain a healthy, innovative Internet.

When an Internet application doesn't work as expected, how can you tell whether the problem is caused by your broadband connection, the application or something else? It can be very difficult for professional network administrators, let alone average Internet users, to answer this sort of question today.

Transparency has always been an essential component of the Internet's success, and Internet users deserve to be well-informed about the performance of their broadband connections. For that to happen, researchers need resources to develop new analytical tools.

That's where M-Lab comes in.

Tools for Users

Many researchers are already developing tools that allow users to test their broadband connections by briefly communicating with a server elsewhere on the Internet. The M-Lab platform supports the development of these measurement tools.

An Open Platform for Researchers

M-Lab assists scientific research by providing widely-distributed servers and ample connectivity for researchers' use. Each tool will be allocated dedicated resources on the M-Lab platform to facilitate accurate measurements. Server-side tools will be openly licensed and operated to allow third-parties to develop their own client-side measurement software.

Better, Open Data for Everyone

All data collected via M-Lab will be made available to the research community to allow researchers to build on a common pool of network measurement data.

M-Lab today and in the future

M-Lab is a collaborative effort led by researchers in partnership with companies and other institutions. Today, M-Lab is only at the beginning of its development. Five tools are currently available, running on 45 servers in 15 locations. In order to achieve our goals, M-Lab depends on the active support of additional partners, and we welcome input from anyone interested in helping the platform flourish.

For more information, read this introduction to Measurement Lab and invitation to the research community to get involved, in the June ACM's Computer Communication Review available here.

If you are an Internet user and want to run the tools, check out our Tools page.

If you are a company, institution or researcher interested in learning more about using and supporting M-Lab, learn more here.

M-Lab in GSoC 2012

In the spirit of collaboration that defines M-Lab, we are acting as an umbrella organization for projects mentored by three groups:

1. The M-Lab core team (our ideas here)
   M-Lab is a collaborative effort aimed at empowering users, researchers, and regulators with good data on scientific performance. M-Lab was founded by Vint Cerf, Stephen Stuart, and a large consortium of academic and industry partners. It operates as a consortium, with many organizations contributing work that furthers M-Lab’s goals. In this spirit we are submitting a single M-Lab application that includes work by a number of contributing organizations. All organizations here have made clear commitments to mentorship, and all projects here have been deemed essential and doable by the M-Lab core team. We’ve listed the individual contributing institutions below.
2. Princeton University's DONAR project (their ideas here; scroll down)
   With the advent of cloud computing and the growth of popular Web services, many networked applications are replicated at multiple geographic locations. Such distributed services face the challenge of server selection — that is, directing an incoming client request to the appropriate server or data center, in the hope of reducing network latency or carefully tuning server loads. M-Lab runs on a federated network of world-wide machines. Thus, optimal server selection is critical to ensure the accuracy and performance of the tools and diagnostics which are built on the M-Lab platform.
   
   DONAR is a distributed system that provides name resolution and server selection for M-Lab experiments. It provides replica registration through a service API and offers client routing to replicas via DNS. DONAR explores many research issues surrounding mutli-replica services, including API design and optimal client-server assignment.
3. UPMC Sorbonne Université's Paris Traceroute project (their ideas here)
   UPMC Sorbonne Universites is France’s leading technical university in science, engineering, and medicine, with over 30,000 students at its campuses in central Paris. It can be thought of as the Sorbonne’s “technical wing.” Its networking group is working on Paris Traceroute, a fundamental upgrade to the Traceroute tool that currently exists in all major operating systems. This technology that could help M-Lab gather and analyze network topology data and that would allow it to document the efficacy of its server placement and data selection process. Documenting this will be crucial as M-Lab encourages bigger and more cautious organizations and governments to rely on its data. M-Lab has had multiple in-person meetings with Paris Traceroute team members, and explicitly trusts them to deliver.
   
   The UPMC team has developed Paris Traceroute and released it as free, open source software. Its impact in the network measurement research community has been felt broadly, as evidenced by well over 100 scientific papers that a casual search for “Paris Traceroute” will reveal. There is a “paris-traceroute” package available at all of the official Debian repositories.