Skip to content
Batfish is a network configuration analysis tool that can find bugs and guarantee the correctness of (planned or current) network configurations. It enables network engineers to rapidly and safely evolve their network, without fear of outages or security breaches.
Branch: master
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
.buildkite Buildkite: add emoji to clarify build (#3385) Mar 13, 2019
.idea
networks Adding network with ISP configurations (#3177) Feb 20, 2019
projects
questions Update grammar doc for Application and IpProtocol specification (#3452) Mar 22, 2019
skylark
tests Interface: store name of OSPF process it belongs to (#3449) Mar 22, 2019
tools JsonSchemaDescription -> Javadoc (#3305) Mar 1, 2019
.codecov.yml
.gitattributes update gitattributes for sln files Jun 26, 2017
.gitignore
BUILD Add bazel BUILD config to batfish in background (#1460) May 22, 2018
CONTRIBUTING.md
LICENSE
README.md Fix typo in README (#3374) Mar 12, 2019
WORKSPACE Bazel: fix formatting, add deterministic rule to fix warning (#3380) Mar 12, 2019
batfish_notebook.png
batfish_video.png

README.md

Got questions, feedback, or feature requests? Join our community on Slack!

codecov

What is Batfish?

Batfish is a network validation tool that provides correctness guarantees for security, reliability, and compliance by analyzing the configuration of network devices. It builds complete models of network behavior from device configurations and finds violations of network policies (built-in, user-defined, and best-practices).

A primary use case for Batfish is to validate configuration changes before deployment (though it can be used to validate deployed configurations as well). Pre-deployment validation is a critical gap in existing network automation workflows. By including Batfish in automation workflows, network engineers can close this gap and ensure that only correct changes are deployed.

Batfish does NOT require direct access to network devices. The core analysis requires only the configuration of network devices. This analysis may be enhanced using additional information from the network such as:

  • BGP routes received from external peers
  • Topology information represented by LLDP/CDP

See www.batfish.org for technical information on how it works. Analytics

What kinds of correctness checks does Batfish support?

The Batfish YouTube channel (subscribe!) and Python notebooks illustrate many checks. Batfish checks span a range of network behaviors.

Configuration Compliance

  • Flag undefined-but-referenced or defined-but-unreferenced structures (e.g., ACLs, route maps)
  • Configuration settings for MTUs, AAA, NTP, logging, etc. match templates
  • Devices can only be accessed using SSHv2 and password is not null

Reliability

  • End-to-end reachability is not impacted for any flow after any single-link or single-device failure
  • Certain services (e.g., DNS) are globally reachable

Security

  • Sensitive services can be reached only from specific subnets or devices
  • Paths between endpoints are as expected (e.g., traverse a firewall, have at least 2 way ECMP, etc...)

Change Analysis

  • End-to-end reachability is identical across the current and a planned configuration
  • Planned ACL or firewall changes are provably correct and causes no collateral damage for other traffic
  • Two configurations, potentially from different vendors, are functionally equivalent

How do I get started?

Getting started with Batfish is easy, just grab the latest allinone Batfish Docker container:

docker pull batfish/allinone

The container has:

Once you have installed the container, the first thing we recommend is walking through the Jupyter notebooks. Each notebook highlights different capabilities of Batfish and shows you how to exercise it. Check-out this README for the detailed list of notebooks.

Running the example notebooks

To run through the example Jupyter notebooks, start the docker container:

docker run -p 8888:8888 batfish/allinone

When this container starts, Jupyter will show a token required for access (e.g. token=abcdef123456...). Make note of this, as you will need it to access the Jupyter server.

Now just open your web-browser and navigate to http://localhost:8888 enter the token in the Password or token: prompt to access the notebooks.

Evaluate your own network configurations

Now that you are familiar with the capabilities of Batfish, you are ready to analyze your network configurations. The first thing to do is create a local data directory. This is a folder on the host machine running the docker container, where Batfish will persist data across container reboots.

mkdir -p data

Stop and restart the container.

docker stop $(docker ps -f "ancestor=batfish/allinone" -q)

docker run -v $(pwd)/data:/data -p 9997:9997 -p 9996:9996 batfish/allinone

This starts the service after mapping the local data folder to the data folder within the container and exposing the TCP ports required by the Batfish service.

Next, you need to install Pybatfish (the Python SDK) in order to interact with the service.

Install Pybatfish

We highly recommend that you install Pybatfish in a Python 3 virtual environment. Details on how to set one up can be found here. Once your virtual environment is setup and activated, upgrade pip and then install pybatfish.

python3 -m pip install --upgrade pip
python3 -m pip install --upgrade git+https://github.com/batfish/pybatfish.git

Now, you are ready to evaluate your own network with Batfish. We encourage you to use Jupyter notebooks as your starting point, but you can use other methods that you are a comfortable with, e.g., an IDE like PyCharm or an interactive Python shell. If you choose to use Jupyter notebooks as your starting point, you need to install Jupyter in your virtual environment. Jupyter documentation can be found here - but the commands below will get you going.

python3 -m pip install jupyter
jupyter notebook

Our notebooks provide a quick start guide for different use cases. Beyond that, the complete documentation is available on readthedocs.

System Requirements for running Batfish

Batfish can be run on any operating system that supports Docker. The containers are actively tested on Mac OS X and Ubuntu 16.04 LTS.

To get started with the example Jupyter notebooks, all you need is a reasonably capable laptop:

  • Dual core CPU
  • 8 GB RAM
  • 256 GB hard-drive

When you transition to running Batfish on your own network, we recommend a server that at least has:

  • Quad-core CPU with 2 threads per CPU
  • 32 GB RAM
  • 256 GB hard-drive

Supported Network Device and Operating System List

Batfish supports configurations for a large and growing set of (physical and virtual) devices, including:

  • Arista
  • Aruba
  • AWS (VPCs, Network ACLs, VPN GW, NAT GW, Internet GW, Security Groups, etc…)
  • Cisco (All Cisco NX-OS, IOS, IOS-XE, IOS-XR and ASA devices)
  • Dell Force10
  • F5 BIG-IP
  • Foundry
  • iptables (on hosts)
  • Juniper (All JunOS platforms: MX, EX, QFX, SRX, T-series, PTX)
  • MRV
  • Palo Alto Networks
  • Quagga / FRR
  • Quanta
  • VyOS

If you'd like support for additional vendors or currently-unsupported configuration features, let us know via Slack or GitHub. We'll try to add support. Or, you can -- we welcome pull requests! :)

License and Dependencies

Batfish is released under The Apache Software License, Version 2.0. All third-party dependencies are compatible with this licensing. A full list of dependencies and their licenses can be generated by running mvn -f projects license:aggregate-add-third-party.

You can’t perform that action at this time.