By Jean-François Grailet (last updated: January 7, 2020)
This specific repository provides resources (source code, measurements, additional Python scripts) only for the very first version of SAGE, written in fall 2017 and published online in early 2018.
This early version of SAGE uses roughly the same ideas as the current tool, but is built using the subnet inference parts of TreeNET ( https://github.com/JefGrailet/treenet ) rather than being built on top of WISE ( https://github.com/JefGrailet/WISE ). It also features a less advanced graph building methodology. To avoid any confusion, it is now considered a β version of the current SAGE, as advertised by the repository name.
This repository is left available for curious researchers and for state-of-the-art reasons: SAGE re-using the subnet inference parts of TreeNET, they can be considered as the most up-to-date version of TreeNET regarding subnet inference.
To get the up-to-date SAGE, please visit the public repository of the same name. The (early) version of SAGE provided here will no longer be updated, and no new dataset produced by it will be published here.
https://github.com/JefGrailet/SAGE
SAGE is a topology discovery tool which aims at discovering an accurate picture of a target network by building a graph of its neighborhoods, i.e., network locations bordered by subnets located at at most one hop away from each other and which usually consist of one router or a mesh of routers (to be discovered with alias resolution). To achieve this goal, SAGE scans for subnets in the target domain and collects (Paris) traceroute measurements for each discovered subnet. Then, it runs various algorithms to aggregate the subnets and see how these aggregates are positioned with respect to each other to build a complete graph of the target network, using a minimal amount of information from each route.
In the process, SAGE uses active probing in order to perform alias resolution on router interfaces discovered by traceroute when it discovers these interfaces are topologically close to each other. By only using a small part of the collected routes and performing alias resolution to discover routing convergence points, SAGE avoids many of the issues one can encounter while interpreting traceroute data. SAGE is also able to infer which subnet implements a point-to-point link between two neighborhoods that are one hop away from each other, and therefore provides a very
complete picture of the topology of a network.
To discover subnets, SAGE re-uses state-of-the-art algorithms from both TreeNET and ExploreNET and re-uses as well the alias resolution methodology of TreeNET to discover routers among the neighborhoods found in the final graph.
As the subnet inference algorithms used in SAGE are designed for IPv4, SAGE is currently only available for IPv4. It also comes as a 32-bit application (written in C/C++) to ensure compatibility with all PlanetLab computers.
Since it needs to be compatible with old environments (e.g. 32-bit machines from the PlanetLab testbed running with Fedora 8), SAGE is written in an old-fashioned C++, i.e., it doesn't take advantage of the changes brought by C++11 and onwards. This said, after its numerous campaigns run from PlanetLab towards all kinds of target networks, it is safe to assume SAGE is unlikely to crash or mismanage memory. It has been, on top of that, been extensively tested with valgrind on a local network.
This repository consists of the following content:
-
Dataset/ provides complete datasets for various Autonomous Systems (or ASes) we measured with
SAGEfrom the PlanetLab testbed. Additionnal sub-folders provide Python scripts to have a better look at the data and validate a specific measurement methodology we used to measure large and/or highly responsive ASes. -
Motivations/ provides some datasets collected with
TreeNETin September 2017 along some Python scripts used to investigate the early intuitions behind the design ofSAGE. -
v1/ provides all the source files of
SAGEv1.0, along some instructions to build and use it.
SAGE was written by Jean-François Grailet, currently Ph. D. student at the University of Liège (Belgium) in the Research Unit in Networking (RUN). Some parts of SAGE re-uses code from TreeNET, which itself re-uses code from ExploreNET. For more details on TreeNET and how it relates to previous software, check the TreeNET public repository.
E-mail address: Jean-Francois.Grailet@uliege.be
Personal website: http://www.run.montefiore.ulg.ac.be/~grailet/
Feel free to send me an e-mail if your encounter problems while compiling or running SAGE. I am also inclined to answer questions regarding the algorithms used in SAGE and to discuss its application in other research projects.