docker compose down looks like a tidy shutdown command. It is, but the flags around it decide whether you walk away with your data intact, with a clean image store, or with a database wiped because you typed -v out of habit. The default behavior Without any flag, down stops and removes: Every container in the project The default network Compose created for the project Anonymous volumes that the stack was using It does not remove: ...
docker compose up streams logs from every service in a single terminal. That works fine for two or three services. With a stack that includes a chatty proxy, a healthcheck loop, or a model server printing tokenizer warnings every second, the signal-to-noise ratio drops fast. attach: false tells Compose to keep the service running but stop streaming its logs to the foreground. Per-service in the Compose file Mark the noisy service: ...
docker compose ps shows the services inside the current project. docker compose ls zooms out: every Compose stack running anywhere on the host, regardless of which directory you happen to be standing in. The basics Run it from any directory: docker compose ls Sample output: NAME STATUS CONFIG FILES api-platform running(4) /home/dev/api-platform/compose.yaml data-pipeline running(2) /home/dev/data-pipeline/compose.yaml old-prototype exited(3) /tmp/prototype/compose.yaml Three columns: the project name, how many services are up, and the path to the Compose file that started it. No need to cd into the project to see what’s running. ...
Two directives that look similar but do completely different things. Confusing expose with ports is a classic way to either break inter-service communication or accidentally publish a database to the public internet. What each one does ports publishes a container port to the host. The outside world (and anything on the host) can reach it. services: web: image: nginx ports: - "8080:80" # host:container expose declares that a container listens on a port. It does not publish anything to the host. The directive is documentation of intent for tooling and humans. ...
Linux isolates containers using kernel namespaces. Sometimes you need the opposite: two containers that can see each other’s processes or share memory. The pid and ipc directives give you that escape hatch. Sharing a PID namespace pid: service:<name> lets a container see and act on processes inside another service: services: app: image: myapp debugger: image: alpine pid: service:app cap_add: - SYS_PTRACE command: sleep infinity The debugger container’s ps, strace, and /proc all reflect app’s processes. Combined with cap_add: SYS_PTRACE, you can attach strace or gdb to a running production-style container without baking debug tools into its image. ...