And this is a kind of a surprise for beginners. The IP address returned by this command is the one assigned to a virtual machine created internally by Docker-compose on Windows to implement containers.

It turned out that under Windows (Windows 10, Docker Toolbox for Windows) to establish IP address that is used on the host machine to communicate with containerized applications one needs to use the docker-machine command: We would like to know the IP address used from the host machine to communicate with containers. Please note that we’re not interested in internal IP addresses of containers used in inter-container communication. Those ports can be easily established based on the output of docker-compose ps command. Each container exposed one port that should be mapped by Docker-compose to a dynamically allocated port accessible from the host machine. Unfortunately it looks that Docker-compose for Linux works a little bit different that Docker-compose for Windows (at least in Docker Toolbox variant).Ĭontainers were configured to use network mode “bridge”, so containerized applications could communicate between each other. The aim was to prepare an infrastructure in an OS-independent way. A program running on a host machine needed to know IP address and port number mapped to a given containerized application. Last time I was playing a little bit with setting up a Docker-compose environment with couple of applications running in separate containers and communicating with them from the host machine.

Windows 10 docker container ip address