docker-run

Run a container interactively

$ docker run -it [image] [bash]

$ docker run -d [image]

$ docker run -p [8080:80] [image]

$ docker run -v [/host/path:/container/path] [image]

$ docker run -e [VAR=value] [image]

$ docker run --rm [image]

$ docker run --name [container_name] [image]

$ docker run -m [512m] [image]

docker run [options] image [command] [args]

docker run creates and starts a new container from an image. It combines docker create and docker start into a single command, providing the primary way to launch containers.
The command pulls the image if not present locally, creates a container from it, and starts execution. Options control resource allocation, networking, storage, and runtime behavior.
Containers are isolated but can connect to host resources through port mapping, volume mounts, and network configuration. The -it flags enable interactive terminal access for debugging and exploration.

-d, --detach

Run container in background.

Interactive with pseudo-TTY.

Publish container port to host.

Bind mount a volume.

Set environment variable.

Assign container name.

Remove container on exit.

Memory limit (e.g., 512m, 1g).

Number of CPUs.

Connect to network.

Restart policy: no, always, unless-stopped, on-failure.

Working directory inside container.

Username or UID.

Override default entrypoint.

Containers run as root by default; use -u for non-root execution. Ports below 1024 require root on the host. Volumes with relative paths are created as Docker volumes, not bind mounts. Container filesystem changes are lost unless volumes are used.

docker run has been the fundamental Docker command since the initial release in 2013 by Solomon Hykes and dotCloud (later Docker, Inc.). The command's design draws from Unix philosophy and LXC container concepts. Options have expanded significantly to support enterprise features, orchestration, and security requirements.

docker(1), docker-build(1), docker-ps(1), docker-exec(1)