Using Docker in Cloud Foundry

Page last updated:

You can activate Docker support so Cloud Foundry can deploy and manage apps based on Docker container images.

For information about Diego, the Cloud Foundry Application Runtime component that manages app containers, see Diego Components and Architecture. For information about how Cloud Foundry developers push apps with Docker images, see Deploying an App with Docker.

Activate Docker image support

By default, apps deployed with the cf push command run in Cloud Foundry containers based on standard filesystems that come bundled with CF. With Docker support activated, Cloud Foundry can also deploy and manage apps based on Docker container images.

To deploy Docker-image-based apps, developers run cf push with the --docker-image option and the location of a Docker image from which to create the containers. For information about how Cloud Foundry developers push apps with Docker images, see Push a Docker Image.

To activate Docker support on a Cloud Foundry deployment, you must:

• Activate the diego_docker feature flag.

• Configure access to any Docker registries that you want to use images from.

Configuring these properties requires admin-level access to the Cloud Foundry deployment, and cannot be done by individual app developers.

Enable and deactivate the diego_docker feature flag

The diego_docker feature flag governs whether a Cloud Foundry deployment supports Docker-image-based containers.

To activate Docker image support, run:

cf enable-feature-flag diego_docker

To deactivate Docker image support, run:

cf disable-feature-flag diego_docker

Deactivating the diego_docker feature flag stops all Docker-image-based apps in your deployment within a few convergence cycles, on the order of a one minute.

Configure Docker registry access

To support Docker-image-based apps, Cloud Foundry needs the ability to access Docker registries using either a Certificate Authority or an IP address allow list. For information about configuring this access, see Installing Certificates on VMs.

Docker image contents

A Docker image consists of a collection of layers. Each layer consists of one or both of the following:

• Raw bits to download and mount. These bits form the file system.

• Metadata that describes commands, users, and environment for the layer. This metadata includes the ENTRYPOINT and CMD directives, and is specified in the Dockerfile.

How Garden-runC creates containers

Diego currently uses Garden-runC to construct Linux containers.

Both Docker and Garden-runC use libraries from the Open Container Initiative (OCI) to build Linux containers. After creation, these containers use name space isolation, or namespaces, and control groups, or cgroups, to isolate processes in containers and limit resource usage. These are common kernel resource isolation features used by all Linux containers.

Before Garden-runC creates a Linux container, it creates a file system that is mounted as the root file system of the container. Garden-runC supports mounting Docker images as the root file systems for the containers it creates.

When creating a container, both Docker and Garden-runC:

• Fetch and cache the individual layers associated with a Docker image.
• Combine and mount the layers as the root file system.

These actions produce a container whose contents exactly match the contents of the associated Docker image.

For earlier versions of Diego using Garden-Linux, see Garden.

How Diego runs and monitors processes

After Garden-runC creates a container, Diego runs and monitors the processes inside of it.

To determine which processes to run, the Cloud Controller fetches and stores the metadata associated with the Docker image. The Cloud Controller uses this metadata to:

• Run the start command as the user specified in the Docker image.

• Instruct Diego and the Gorouter to route traffic to the lowest-numbered port exposed in the Docker image, or port 8080 if the Dockerfile does not explicitly expose a listen port.

For more information about Cloud Controller, see Cloud Controller.

For more information about Gorouter and the routing tier, see Cloud Foundry Routing Architecture.

For more information about exposed ports in Docker images, see the Expose section of the Dockerfile reference topic in the Docker documentation.

When you launch an app on Diego, the Cloud Controller honors any user specified overrides. For example, a custom start command or custom environment variables.

Docker security concerns in a multi-tenant environment

The attack surface area for a Docker-image-based container that runs on Diego remains somewhat higher than that of a buildpack app because Docker allows you to fully specify the contents of your root file systems. A buildpack app runs on a trusted root filesystem.

Garden-runC provides features that allow the platform to run Docker images more securely in a multitenant context. In particular, Cloud Foundry uses the “user-namespacing” feature found on modern Linux kernels to ensure that users cannot gain escalated privileges on the host even if they escalate privileges within a container.

The Cloud Controller always runs Docker containers on Diego with user namespaces enabled. This security restriction prevents certain features, for example, the ability to mount FuseFS devices, from working in Docker containers. Docker apps can use fuse mounts through volume services, but they cannot directly mount fuse devices from within the container.

For more information about volume services, see Using an External File System (Volume Services).

To mitigate security concerns, Cloud Foundry run only trusted Docker containers on the platform. By default, the Cloud Controller does not allow Docker based apps to run on the platform.