Machine concepts and getting helpEstimated reading time: 4 minutes
Docker Machine allows you to provision Docker machines in a variety of environments, including virtual machines that reside on your local system, on cloud providers, or on bare metal servers (physical computers). Docker Machine creates a Docker host, and you use the Docker Engine client as needed to build images and create containers on the host.
Drivers for creating machines
To create a virtual machine, you supply Docker Machine with the name of the driver you want use. The driver determines where the virtual machine is created. For example, on a local Mac or Windows system, the driver is typically Oracle VirtualBox. For provisioning physical machines, a generic driver is provided. For cloud providers, Docker Machine supports drivers such as AWS, Microsoft Azure, Digital Ocean, and many more. The Docker Machine reference includes a complete list of supported drivers.
Default base operating systems for local and cloud hosts
Since Docker runs on Linux, each VM that Docker Machine provisions relies on a base operating system. For convenience, there are default base operating systems. For the Oracle Virtual Box driver, this base operating system is boot2docker. For drivers used to connect to cloud providers, the base operating system is Ubuntu 12.04+. You can change this default when you create a machine. The Docker Machine reference includes a complete list of supported operating systems.
IP addresses for Docker hosts
For each machine you create, the Docker host address is the IP address of the
Linux VM. This address is assigned by the
docker-machine create subcommand.
You use the
docker-machine ls command to list the machines you have created.
docker-machine ip <machine-name> command returns a specific host’s IP
Configuring CLI environment variables for a Docker host
Before you can run a
docker command on a machine, you need to configure your
command-line to point to that machine. The
docker-machine env <machine-name>
subcommand outputs the configuration command you should use.
For a complete list of
docker-machine subcommands, see the
Docker Machine subcommand reference.
Custom root Certificate Authority for Registry
Users using their own Docker Registry will experience
x509: certificate signed by unknown authority
error messages if their registry is signed by custom root Certificate Authority and it is
not registered with Docker Engine. As discussed in the
Docker Engine documentation
certificates should be placed at
hostname is your Registry server’s hostname.
docker-machine scp certfile default:ca.crt docker-machine ssh default sudo mv ~/ca.crt /etc/docker/certs.d/hostname/ca.crt exit docker-machine restart
Provisioning a host is a complex matter that can fail for a lot of reasons. Your workstation may have a wide variety of shell, network configuration, VPN, proxy or firewall issues. There are also reasons from the other end of the chain: your cloud provider or the network in between.
docker-machine be as stable as possible, we added a monitoring of
crashes whenever you try to
upgrade a host. This will send, over
HTTPS, to Bugsnag some information about your
docker-machine version, build,
OS, ARCH, the path to your current shell and, the history of the last command as
you could see it with a
--debug option. This data is sent to help us pinpoint
recurring issues with
docker-machine and will only be transmitted in the case
of a crash of
If you wish to opt out of error reporting, you can create a
file in your
$HOME/.docker/machine directory, and Docker Machine will disable
this behavior. e.g.:
$ mkdir -p ~/.docker/machine && touch ~/.docker/machine/no-error-report
Leaving the file empty is fine – Docker Machine just checks for its presence.
Docker Machine is still in its infancy and under active development. If you need help, would like to contribute, or simply want to talk about the project with like-minded individuals, we have a number of open channels for communication.
- To report bugs or file feature requests: please use the issue tracker on Github.
- To talk about the project with people in real time: please join the
#docker-machinechannel on IRC.
- To contribute code or documentation changes: please submit a pull request on Github.
For more information and resources, please visit our help page.