w3af inside docker

Using w3af inside docker should be transparent for most use cases, this page documents the use cases which are complex to solve when docker is added to the mix.

Ports and services

Some w3af plugins, such as crawl.spider_man and audit.rfi start proxy HTTP services. In order to access these services the plugins need to be configured to listen on 0.0.0.0 and the port needs to be made accessible to the host using the -p parameter in the helper script (ie. extras/docker/scripts/w3af_console_docker)

Take a look at this commit for more information about exposing ports.

Sharing data with the container

When starting w3af using the w3af_console_docker or w3af_gui_docker commands the docker containers are started with two volumes which are mapped to your home directory:

  • ~/.w3af/ from your host is mapped to /root/.w3af/ in the container. This directory is mostly used by w3af to store scan profiles and internal data.
  • ~/w3af-shared from your host is mapped to /root/w3af-shared in the container. Use this directory to save your scan results and provide input files to w3af.

Debugging the container

The container runs a SSH daemon, which can be used to both run the w3af_console and w3af_gui. To connect to a running container use root as username and w3af as password. Usually you don’t need to worry about this, since the helper scripts will connect to the container for you.

Another way to debug the container is to run the script with the -d flag:

$ sudo ./w3af_console_docker -d
root@a01aa9631945:~#

Note

WARNING: Don’t bind w3af’s docker image to a public IP address unless you really know what you’re doing! Anyone will be able to SSH into the docker image using the hard-coded SSH keys!