Glasnost: Test if your ISP is shaping your traffic

* Home * Glasnost Tests * Create your own test * Run your own server * Results * Contact *


The goal of the Glasnost project is to make ISPs' traffic shaping policies transparent to their customers. To this end, we designed Glasnost tests that enable you to check whether traffic from your applications is being rate-limited (i.e., throttled) or blocked.

Glasnost tests work by measuring and comparing the performance of different application flows between your host and our measurement servers. The tests can detect traffic shaping in both upstream and downstream directions separately. The tests can also detect whether application flows are shaped based on their port numbers or their packets' payload. For more details on how Glasnost tests work, please read our NSDI 2010 paper.

We configured our tests to be conservative when declaring the presence of shaping, i.e., passing our tests does not necessarily mean that there is no throttling occurring on your link.

Select a Glasnost test to run
P2P apps   Standard apps   Video-on-Demand
  Email (POP)
Email (IMAP4)
HTTP transfer
SSH transfer
Usenet (NNTP) NEW!
  Flash video (e.g., YouTube)

  • Each Glasnost test takes approximately 8 minutes
  • Note to all users: To allow accurate measurements you should stop any large downloads that might run in the background.

M-Lab Glasnost makes use of the Measurement Lab (M-Lab) research platform.
To learn what information our tool collects, please go here.


Who are we?

We are researchers at the Max Planck Institute for Software Systems. Our research focuses on characterizing residential broadband networks and understanding their implications for the designers of future protocols and applications. In case you have questions about this tool or our research, please visit our network transparency project webpage or contact us via e-mail: broadband @at@ mpi-sws org

Faculty   Students   Alumni   Programmer
* Krishna P. Gummadi
* Ratul Mahajan (Microsoft Research)
* Stefan Saroiu (Microsoft Research)
  * Marcel Dischinger
* Massimiliano Marcon
  * Saikat Guha (Microsoft Research)
* Andreas Haeberlen (University of Pennsylvania)
* Alan Mislove (North Eastern University)
  * Jeff Hoye