refactor: Remove STANDARD and VLLM_NIXL choices from build/run (#148)

README.md

@@ -48,14 +48,12 @@ Dynamo development and examples are container based.
 You can build the Dynamo container using the build scripts
 in `container/` (or directly with `docker build`).
 
-We provide 3 types of builds:
+We provide 2 types of builds:
 
-1. `STANDARD` which includes our default set of backends (onnx, openvino...)
+1. `VLLM` which includes our VLLM backend using new NIXL communication library.
 2. `TENSORRTLLM` which includes our TRT-LLM backend
-3. `VLLM` which includes our VLLM backend using NCCL communication library.
-4. `VLLM_NIXL` which includes our VLLM backend using new NIXL communication library.
 
-For example, if you want to build a container for the `STANDARD` backends you can run
+For example, if you want to build a container for the `VLLM` backend you can run
 
 <!--pytest.mark.skip-->
 ```bash

container/Dockerfile

-# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
-# SPDX-License-Identifier: Apache-2.0
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-# http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-
-ARG BASE_IMAGE="nvcr.io/nvidia/tritonserver"
-ARG BASE_IMAGE_TAG="25.01-py3"
-
-FROM ${BASE_IMAGE}:${BASE_IMAGE_TAG} AS dynamo
-
-# TODO: non root user by default
-
-USER root
-
-# TODO: separate dev from runtime dependendcies
-
-# Rust build/dev dependencies
-RUN apt-get update && \
-    apt-get install --no-install-recommends --yes  gdb protobuf-compiler cmake libssl-dev pkg-config
-
-RUN curl https://sh.rustup.rs -sSf | bash -s -- -y
-ENV PATH="/root/.cargo/bin:${PATH}"
-RUN rustup toolchain install 1.85.0-x86_64-unknown-linux-gnu
-
-# Install OpenAI-compatible frontend and its dependencies from triton server
-# repository. These are used to have a consistent interface, schema, and FastAPI
-# app between Triton Core and Dynamo implementations.
-ARG OPENAI_SERVER_TAG="r25.01"
-RUN mkdir -p /opt/tritonserver/python && \
-    cd /opt/tritonserver/python && \
-    rm -rf openai && \
-    git clone -b ${OPENAI_SERVER_TAG} --single-branch https://github.com/triton-inference-server/server.git && \
-    cd server && \
-    git checkout ${SERVER_OPENAI_COMMIT} && \
-    cd .. && \
-    mv server/python/openai openai && \
-    chown -R root:root openai && \
-    chmod 755 openai && \
-    chmod -R go-w openai && \
-    rm -rf server && \
-    python3 -m pip install -r openai/requirements.txt
-
-# Common dependencies
-RUN --mount=type=bind,source=./container/deps/requirements.txt,target=/tmp/requirements.txt \
-    pip install --timeout=2000 --requirement /tmp/requirements.txt
-RUN --mount=type=bind,source=./container/deps/requirements.nats.txt,target=/tmp/requirements.txt \
-    pip install --timeout=2000 --requirement /tmp/requirements.txt
-RUN --mount=type=bind,source=./container/deps/requirements.test.txt,target=/tmp/requirements.txt \
-    pip install --timeout=2000 --requirement /tmp/requirements.txt
-
-# Finish pyright install
-RUN pyright --help > /dev/null 2>&1
-
-# In Process Python API Install
-RUN find /opt/tritonserver/python -maxdepth 1 -type f -name \
-    "tritonserver-*.whl" | xargs -I {} pip3 install --force-reinstall --upgrade {}[all]
-
-# GENAI Perf Install
-# TODO: Move to tag when fix for genai-perf will be released
-ARG GENAI_PERF_TAG="25d0188713adc47868d6b3f22426375237a90529"
-RUN pip install "git+https://github.com/triton-inference-server/perf_analyzer.git@${GENAI_PERF_TAG}#subdirectory=genai-perf"
-
-# Backend & Framework Specific Installation
-ARG FRAMEWORK="STANDARD"
-RUN --mount=type=bind,source=./container/deps/requirements.standard.txt,target=/tmp/requirements.txt \
-    if [[ "$FRAMEWORK" == "STANDARD" ]] ; then pip install --timeout=2000 -r /tmp/requirements.txt ; fi
-
-
-# TODO set VLLM Version
-# ENV VLLM_VERSION
-ARG VLLM_FRAMEWORK
-# DEFAULT VLLM VARIABLES
-# ENV VLLM_ATTENTION_BACKEND=${VLLM_FRAMEWORK:+FLASHINFER}
-ENV VLLM_WORKER_MULTIPROC_METHOD=${VLLM_FRAMEWORK:+spawn}
-ENV VLLM_TORCH_HOST=${VLLM_FRAMEWORK:+localhost}
-ENV VLLM_TORCH_PORT=${VLLM_FRAMEWORK:+36183}
-ENV VLLM_DATA_PLANE_BACKEND=${VLLM_FRAMEWORK:+nccl}
-ENV VLLM_BASELINE_WORKERS=${VLLM_FRAMEWORK:+0}
-ENV VLLM_CONTEXT_WORKERS=${VLLM_FRAMEWORK:+1}
-ENV VLLM_GENERATE_WORKERS=${VLLM_FRAMEWORK:+1}
-ENV VLLM_BASELINE_TP_SIZE=${VLLM_FRAMEWORK:+1}
-ENV VLLM_CONTEXT_TP_SIZE=${VLLM_FRAMEWORK:+1}
-ENV VLLM_GENERATE_TP_SIZE=${VLLM_FRAMEWORK:+1}
-ENV VLLM_KV_CAPI_PATH="/opt/dynamo/bindings/lib/libdynamo_llm_capi.so"
-ENV PYTHONUNBUFFERED=1
-
-# Install NATS - pointing toward NATS github instead of binaries.nats.dev due to server instability
-RUN wget https://github.com/nats-io/nats-server/releases/download/v2.10.24/nats-server-v2.10.24-amd64.deb && dpkg -i nats-server-v2.10.24-amd64.deb && rm nats-server-v2.10.24-amd64.deb
-
-# etcd
-ENV ETCD_VERSION="v3.5.18"
-RUN wget https://github.com/etcd-io/etcd/releases/download/$ETCD_VERSION/etcd-$ETCD_VERSION-linux-amd64.tar.gz -O /tmp/etcd.tar.gz && \
-    mkdir -p /usr/local/bin/etcd && \
-    tar -xvf /tmp/etcd.tar.gz -C /usr/local/bin/etcd --strip-components=1 && \
-    rm /tmp/etcd.tar.gz
-ENV PATH=/usr/local/bin/etcd/:$PATH
-
-# Enable Git operations in the /workspace directory.
-RUN printf "[safe]\n      directory=/workspace\n" > /root/.gitconfig
-
-# emacs docker-tramp requires /bin/sh to be linked to bash to operate correctly
-RUN ln -sf /bin/bash /bin/sh
-
-# Install NGINX
-RUN apt-get install nginx -y
-RUN rm -rf /etc/nginx/sites-enabled/default
-
-# Install demo utils
-RUN apt-get install nvtop -y
-RUN apt-get install tmux -y
-
-# Working directory
-WORKDIR /workspace
-
-# Copy Python wheel configuration files
-COPY pyproject.toml /workspace/
-COPY README.md /workspace/
-COPY LICENSE /workspace/
-COPY Cargo.toml /workspace/
-COPY Cargo.lock /workspace/
-COPY rust-toolchain.toml /workspace/
-
-ARG CARGO_BUILD_JOBS
-
-ENV CARGO_TARGET_DIR=/workspace/target
-# Build Rust
-COPY lib/ /workspace/lib/
-COPY components /workspace/components
-COPY launch /workspace/launch
-
-RUN cargo build --release --locked --features mistralrs,sglang,vllm,python && cargo doc --no-deps && \
-    cp target/release/dynamo-run /usr/local/bin/ && \
-    cp target/release/llmctl /usr/local/bin/
-
-COPY deploy/dynamo/sdk /workspace/deploy/dynamo/sdk
-
-# Install uv, create virtualenv for general use, and build dynamo wheel
-COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvx /bin/
-RUN mkdir /opt/dynamo && \
-    uv venv /opt/dynamo/venv --python 3.12 && \
-    source /opt/dynamo/venv/bin/activate && \
-    uv build --wheel --out-dir /workspace/dist && \
-    uv pip install /workspace/dist/ai_dynamo*cp312*.whl && \
-    cd /workspace/deploy/dynamo/sdk && \
-    uv build --wheel --out-dir /workspace/dist && \
-    uv pip install /workspace/dist/ai_dynamo_sdk*any.whl
-
-# Package the bindings
-RUN mkdir -p /opt/dynamo/bindings/wheels && \
-    mkdir /opt/dynamo/bindings/lib && \
-    cp dist/ai_dynamo*cp312*.whl /opt/dynamo/bindings/wheels/. && \
-    cp target/release/libdynamo_llm_capi.so /opt/dynamo/bindings/lib/. && \
-    cp -r lib/bindings/c/include /opt/dynamo/bindings/.
-
-# Install dynamo.runtime and dynamo.llm wheels globally in container for tests that
-# currently run without virtual environment activated.
-# TODO: In future, we may use a virtualenv for everything and remove this.
-RUN cd /opt/dynamo/bindings/wheels && \
-    pip install ai_dynamo*cp312*.whl  && \
-    pip install /workspace/dist/ai_dynamo_sdk*any.whl
-
-# Copy everything in after ginstall steps to avoid re-running build/install
-# commands on unrelated changes in other dirs.
-COPY . /workspace
-
-# Enable system UCX
-ENV RAPIDS_LIBUCX_PREFER_SYSTEM_LIBRARY=true
-
-# Command and Entrypoint
-CMD []
-ENTRYPOINT ["/opt/nvidia/nvidia_entrypoint.sh"]

container/build.sh

@@ -43,7 +43,7 @@ PYTHON_PACKAGE_VERSION=${current_tag:-$latest_tag.dev+$commit_id}
 # dependencies are specified in the /container/deps folder and
 # installed within framework specific sections of the Dockerfile.
 
-declare -A FRAMEWORKS=(["STANDARD"]=1 ["TENSORRTLLM"]=2 ["VLLM"]=3 ["VLLM_NIXL"]=4)
+declare -A FRAMEWORKS=(["VLLM"]=1 ["TENSORRTLLM"]=2)
 DEFAULT_FRAMEWORK=VLLM
 
 SOURCE_DIR=$(dirname "$(readlink -f "$0")")
@@ -51,12 +51,8 @@ DOCKERFILE=${SOURCE_DIR}/Dockerfile
 BUILD_CONTEXT=$(dirname "$(readlink -f "$SOURCE_DIR")")
 
 # Base Images
-
-STANDARD_BASE_VERSION=25.01
-STANDARD_BASE_IMAGE=nvcr.io/nvidia/tritonserver
-STANDARD_BASE_IMAGE_TAG=${STANDARD_BASE_VERSION}-py3
-
 TENSORRTLLM_BASE_VERSION=25.01
+# FIXME: Need a public image for public consumption
 TENSORRTLLM_BASE_IMAGE="gitlab-master.nvidia.com:5005/dl/dgx/tritonserver/tensorrt-llm/amd64"
 TENSORRTLLM_BASE_IMAGE_TAG=krish-fix-trtllm-build.23766174
 TENSORRTLLM_PIP_WHEEL_PATH=""
@@ -194,10 +190,6 @@ get_options() {
         FRAMEWORK=$DEFAULT_FRAMEWORK
     fi
 
-    if [[ ${FRAMEWORK^^} == "VLLM_NIXL" ]]; then
-	FRAMEWORK="VLLM"
-    fi
-
     if [ ! -z "$FRAMEWORK" ]; then
         FRAMEWORK=${FRAMEWORK^^}
 

container/run.sh

@@ -22,7 +22,7 @@ RUN_PREFIX=
 # dependencies are specified in the /container/deps folder and
 # installed within framework specific sections of the Dockerfile.
 
-declare -A FRAMEWORKS=(["STANDARD"]=1 ["TENSORRTLLM"]=2 ["VLLM"]=3 ["VLLM_NIXL"]=4)
+declare -A FRAMEWORKS=(["VLLM"]=1 ["TENSORRTLLM"]=2)
 DEFAULT_FRAMEWORK=VLLM
 
 SOURCE_DIR=$(dirname "$(readlink -f "$0")")
@@ -170,10 +170,6 @@ get_options() {
 	FRAMEWORK=$DEFAULT_FRAMEWORK
     fi
 
-    if [[ ${FRAMEWORK^^} == "VLLM_NIXL" ]]; then
-	FRAMEWORK="VLLM"
-    fi
-
     if [ ! -z "$FRAMEWORK" ]; then
 	FRAMEWORK=${FRAMEWORK^^}
 	if [[ ! -n "${FRAMEWORKS[$FRAMEWORK]}" ]]; then