Offline IP address to Autonomous System Number lookup module.
pyasn is a Python extension module that enables very fast IP address to Autonomous System Number lookups. Current state and Historical lookups can be done, based on the MRT/RIB BGP archive used as input.
pyasn is different from other ASN lookup tools in that it provides offline and historical lookups. It provides utility scripts for users to build their own lookup databases based on any MRT/RIB archive. This makes pyasn much faster than online dig/whois/json lookups.
The module is written in C and Python, and cross-compiles on Linux and Windows. Underneath, it uses a radix tree data structure for storage of IP addresses. In the current version, it borrows code from py-radix to support both IPV4 and IPV6 network prefixes. The current release is a beta. Compared to the previous version, it provides support for Python 2 and 3; adds new functionality, performance improvements, and unit-tests.
pyasn is developed and maintained by researchers at the Economics of Cybersecurity research group at Delft University of Technology (http://econsec.tbm.tudlft.nl). The package is used on an almost daily basis and bugs are fixed pretty quickly. The package is largely developed and maintained by Hadi Asghari and Arman Noroozian. Please report any bugs via GitHub (https://github.com/hadiasghari/pyasn) or email the developers.
Installation is a breeze via pip:
pip install pyasn --pre
Or with the standard Python:
python setup.py build python setup.py install --record log
You will need to have pip, setuptools and build essentials installed if you build the package manually. On Ubuntu/Debian you can get them using the following command:
sudo apt-get install python-pip python-dev build-essential
Building the C module on Windows, using either pip or from source, requires Microsoft Visual C++ to be installed. pyasn has been tested using Visual C++ Express 2010, available freely from Microsoft’s website, on both the official Python 3.4 release and Miniconda3. Other versions of Python, Visual Studio, and Cygwin could also work with minor modifications.
We plan to release pyasn packages to major Linux repositories once it is out of beta.
A simple example that demonstrates most of the features:
import pyasn # Initialize module and load IP to ASN database # the sample database can be downloaded or built - see below asndb = pyasn.pyasn('ipasn_20140513.dat') asndb.lookup('126.96.36.199') # should return: (15169, '188.8.131.52/24'), the origin AS, and the BGP prefix it matches asndb.get_as_prefixes(1128) # returns ['184.108.40.206/16', '220.127.116.11/16', '18.104.22.168/16'], TU-Delft prefixes
IPASN Data Files
IPASN data files are a long list of prefixes used to lookup AS number for IPs. An excerpt from such a file looks like this:
; IP-ASN32-DAT file ; Original file : <Path to a rib file> ; Converted on : Tue May 13 22:03:05 2014 ; CIDRs : 512490 ; 22.214.171.124/24 15169 126.96.36.199/17 9737 188.8.131.52/18 9737 184.108.40.206/19 9737 220.127.116.11/24 23969 ...
IPASN data files can be created by downloading MRT/RIB BGP archives from Routeviews (or similar sources), and parsing them using provided scripts that tail the BGP AS-Path. This can be done simply as follows:
pyasn_util_download.py --latest pyasn_util_convert.py --single <Downloaded RIB File> <ipasn_db_file_name>
NOTE: These scripts are by default installed to /usr/local/bin and can be executed directly. If you installed the package to a user directory, these scripts will not be on the path and you will have to invoke them by navigating to the folder in which they have been copied (e.g. ~/.local/bin).
We also provide download links to a large number of previously generated IPASN data files. These are based on weekly snapshots of the Routeviews data from 2005-2015, accessible here: http://data.3tu.nl/repository/uuid:d4d23b8e-2077-4592-8b47-cb476ad16e12
New in v1.6: To save disk space, you can gzip IPASN data files. The load time will be slighlty longer.
Initial loading of a IPASN data file is the most heavy operation of the package. For fast lookups using multiple IPASN data files, for instance for historical lookups on multiple dates, we recommend caching of loaded data files for better performance.
You can remove pyasn as follows:
pip uninstall pyasn
If you built and installed the package your self use the recorded log to remove the installed files.
Removing PyASN version 1.2: pyasn v1.5/1.6 and v1.2 can be installed side by side (due to lower-cased package name). To avoid mistakes, you can uninstall the old PyASN by deleting the following files from your Python installation:
The main portions of the directory tree are as follows:
. ├── pyasn/__init__.py # Python code of the main pyasn module ├── pyasn/pyasn_radix.c # C extension code (Python RADIX module with bulk load) ├── pyasn/_radix/* # C extension code (Based on original RADIX code from MRTd) ├── pyasn/mrtx.py # python module used to convert MRT files to pyasn DB files ├── pyasn-utils/*.py # Scripts to download & convert BGP MRT dumps to IPASN databases ├── data/ # Test Resources and some sample DBs to use ├── tests/ # Tests └── setup.py # Standard setup.py for installation/testing/etc.
Testing pyasn Sources
A limited number of unit tests are provided in the tests/ directory when downlading the sources. They can be run with the following command:
python setup.py test
This beta release has been tested under python version 2.6, 2.7, 3.3, 3.4 and 3.5. We appreciate contributions towards testing pyasn!
New in v1.6: pyasn_util_convert.py offers a ‘–dump-screen’ option which shows the MRT/RIB archive contents and the chosen origin-AS.
License & Acknowledgments
pyasn is licensed under the MIT license.
It extends code from py-radix (Michael J. Schultz and Damien Miller), and improves upon it in several ways, for instance in lowering memory usage and adding bulk prefix/origin load. The underlying radix tree implementation is taken (and modified) from MRTd. These are all subject to their respective licenses. Please see the LICENSE file for details.
Thanks to Dr. Chris Lee (of Shadowserver) for proposing the use of radix trees.
A handful of GitHub developers have contributed features and bug fixes to the latest releases. Many thanks to all of them.