vllm/docs/contributing/dockerfile/dockerfile.md

Dockerfile

We provide a docker/Dockerfile to construct the image for running an OpenAI compatible server with vLLM. More information about deploying with Docker can be found here.

Below is a visual representation of the multi-stage Dockerfile. The build graph contains the following nodes:

• All build stages
• The default build target (highlighted in grey)
• External images (with dashed borders)

The edges of the build graph represent:

• FROM ... dependencies (with a solid line and a full arrow head)

• COPY --from=... dependencies (with a dashed line and an empty arrow head)

• RUN --mount=(.\*)from=... dependencies (with a dotted line and an empty diamond arrow head)

The test-deps stage branches from vllm-runtime-base so Git and the test requirements remain cached independently of per-commit vLLM wheels and source.

The vllm-runtime-base stage explicitly installs the current Ubuntu ca-certificates package so clients using the system trust store receive certificate removals when the image is rebuilt. This complements the Python certifi bundle used by other HTTP clients.

The extensions-build stage can also produce an optional source-built Triton wheel. vllm-openai-base installs that wheel after its other Python dependencies so dependency resolution cannot restore an older Triton version.

With BUILD_NIXL=true, extensions-build also builds NIXL, including its NIXL EP extension for the installed PyTorch, with tools/build_nixl_with_torch215_for_rubin.py. vllm-openai-base then replaces the NIXL packages installed from the KV-connector requirements with these wheels.

The AWS EFA images are built separately by docker/Dockerfile.efa on top of a vllm-openai image built with INSTALL_KV_CONNECTORS=true. It adds a minimal AWS EFA userspace install (libfabric, aws-ofi-nccl and hwloc), the EFA Mooncake wheel (x86_64 only), and DeepEP built for NCCL GIN over EFA. Kernel modules and host limits remain host-managed.

docker build -f docker/Dockerfile.efa \
    --build-arg BASE_IMAGE=vllm/vllm-openai:latest -t vllm:openai-efa docker/

The image does not change vLLM's Mooncake protocol default (rdma). To use EFA, set it per connector:

• MooncakeConnector: "mooncake_protocol": "efa" in kv_connector_extra_config.
• MooncakeStoreConnector: "protocol": "efa" in the JSON file at MOONCAKE_CONFIG_PATH.

The EFA Mooncake wheel is x86_64-only, so use tcp on arm64.

{ align="center" alt="query" width="100%" }

Made using: https://github.com/patrickhoefler/dockerfilegraph

Commands to regenerate the build graph (make sure to run it from the `root` directory of the vLLM repository where the dockerfile is present):

dockerfilegraph \
  -o png \
  --concentrate \
  --legend \
  --dpi 200 \
  --max-label-length 50 \
  --filename docker/Dockerfile

or in case you want to run it directly with the docker image:

docker run \
   --rm \
   --user "$(id -u):$(id -g)" \
   --workdir /workspace \
   --volume "$(pwd)":/workspace \
   ghcr.io/patrickhoefler/dockerfilegraph:alpine \
   --output png \
   --dpi 200 \
   --max-label-length 50 \
   --filename docker/Dockerfile \
   --concentrate \
   --legend

(To run it for a different file, you can pass in a different argument to the flag --filename.)