[None][infra] install mooncake in docker images (#8447)

Signed-off-by: Bo Deng <deemod@nvidia.com>
Signed-off-by: zhengd-nv <200704041+zhengd-nv@users.noreply.github.com>
Signed-off-by: Zheng Duan <200704041+zhengd-nv@users.noreply.github.com>
Co-authored-by: zhengd-nv <200704041+zhengd-nv@users.noreply.github.com>

Author: bo-nv

ATTRIBUTIONS-CPP-aarch64.md

@@ -14889,6 +14889,24 @@ Chen, Tianqi
 
 ```
 
+## Mooncake
+
+- **Repository URL**: https://github.com/kvcache-ai/Mooncake
+- **License URL**: https://github.com/kvcache-ai/Mooncake/blob/main/LICENSE-APACHE
+- **License name**: Apache 2.0
+
+### Authors
+
+© Copyright 2025, Mooncake Team.
+Copyright (c) Meta Platforms, Inc. and affiliates.
+Copyright 2024 KVCache.AI
+Ruoyu Qin
+Zheming Li
+Weiran He
+Mingxing Zhang
+Yongwei Wu
+Weimin Zheng
+Xinran Xu
 ## flashinfer
 
 ### License Text

ATTRIBUTIONS-CPP-x86_64.md

@@ -14697,6 +14697,24 @@ Chen, Tianqi
 
 ```
 
+## Mooncake
+
+- **Repository URL**: https://github.com/kvcache-ai/Mooncake
+- **License URL**: https://github.com/kvcache-ai/Mooncake/blob/main/LICENSE-APACHE
+- **License name**: Apache 2.0
+
+### Authors
+
+© Copyright 2025, Mooncake Team.
+Copyright (c) Meta Platforms, Inc. and affiliates.
+Copyright 2024 KVCache.AI
+Ruoyu Qin
+Zheming Li
+Weiran He
+Mingxing Zhang
+Yongwei Wu
+Weimin Zheng
+Xinran Xu
 ## flashinfer
 
 ### License Text

docker/Dockerfile.multi

@@ -44,6 +44,7 @@ COPY docker/common/install.sh \
      docker/common/install_ucx.sh \
      docker/common/install_nixl.sh \
      docker/common/install_etcd.sh \
+     docker/common/install_mooncake.sh \
      ./
 
 RUN GITHUB_MIRROR=${GITHUB_MIRROR} \
@@ -103,6 +104,13 @@ COPY docker/common/install_triton.sh \
 
 RUN GITHUB_MIRROR=${GITHUB_MIRROR} bash ./install_triton.sh && rm install_triton.sh
 
+# Install Mooncake, after triton handles boost requirement
+RUN if [ -f /etc/redhat-release ]; then \
+        echo "Rocky8 detected, skipping mooncake installation"; \
+    else \
+        bash ./install_mooncake.sh; \
+    fi && rm install_mooncake.sh
+
 FROM ${DEVEL_IMAGE} AS wheel
 WORKDIR /src/tensorrt_llm
 COPY benchmarks benchmarks

docker/common/install_mooncake.sh

@@ -0,0 +1,51 @@
+#!/bin/bash
+set -ex
+
+MOONCAKE_VERSION="v0.3.6.post1"
+MOONCAKE_REPO="https://github.com/kvcache-ai/Mooncake.git"
+MOONCAKE_INSTALL_PATH="/usr/local/Mooncake"
+
+apt-get update
+
+# https://kvcache-ai.github.io/Mooncake/getting_started/build.html
+# libboost-all-dev is removed because it will install a duplicated MPI library
+# triton also installed boost so the requirement is already met
+apt-get install -y --no-install-recommends \
+    build-essential \
+    libibverbs-dev \
+    libgoogle-glog-dev \
+    libgtest-dev \
+    libjsoncpp-dev \
+    libnuma-dev \
+    libunwind-dev \
+    libssl-dev \
+    libyaml-cpp-dev \
+    libcurl4-openssl-dev \
+    libhiredis-dev \
+    pkg-config \
+    patchelf
+
+mkdir -p /third-party-source
+
+git clone --depth 1 https://github.com/alibaba/yalantinglibs.git
+tar -czf /third-party-source/yalantinglibs.tar.gz yalantinglibs
+cd yalantinglibs
+mkdir build && cd build
+cmake .. -DBUILD_EXAMPLES=OFF -DBUILD_BENCHMARK=OFF -DBUILD_UNIT_TESTS=OFF
+make -j
+make install
+cd ../..
+rm -rf yalantinglibs
+
+git clone --depth 1 -b ${MOONCAKE_VERSION} ${MOONCAKE_REPO}
+tar -czf /third-party-source/Mooncake-${MOONCAKE_VERSION}.tar.gz Mooncake
+cd Mooncake
+git submodule update --init --recursive --depth 1
+mkdir build && cd build
+cmake .. -DUSE_CUDA=ON -DBUILD_SHARED_LIBS=ON -DCMAKE_INSTALL_PREFIX=${MOONCAKE_INSTALL_PATH}
+make -j
+make install
+cd ../..
+rm -rf Mooncake
+
+echo "export LD_LIBRARY_PATH=${MOONCAKE_INSTALL_PATH}/lib:\$LD_LIBRARY_PATH" >> "${ENV}"

jenkins/current_image_tags.properties

@@ -13,7 +13,7 @@
 #     images are adopted from PostMerge pipelines, the abbreviated commit hash is used instead.
 IMAGE_NAME=urm.nvidia.com/sw-tensorrt-docker/tensorrt-llm
 
-LLM_DOCKER_IMAGE=urm.nvidia.com/sw-tensorrt-docker/tensorrt-llm:pytorch-25.10-py3-x86_64-ubuntu24.04-trt10.13.3.9-skip-tritondevel-202511021230-8838
-LLM_SBSA_DOCKER_IMAGE=urm.nvidia.com/sw-tensorrt-docker/tensorrt-llm:pytorch-25.10-py3-aarch64-ubuntu24.04-trt10.13.3.9-skip-tritondevel-202511021230-8838
-LLM_ROCKYLINUX8_PY310_DOCKER_IMAGE=urm.nvidia.com/sw-tensorrt-docker/tensorrt-llm:cuda-13.0.2-devel-rocky8-x86_64-rocky8-py310-trt10.13.3.9-skip-tritondevel-202511021230-8838
-LLM_ROCKYLINUX8_PY312_DOCKER_IMAGE=urm.nvidia.com/sw-tensorrt-docker/tensorrt-llm:cuda-13.0.2-devel-rocky8-x86_64-rocky8-py312-trt10.13.3.9-skip-tritondevel-202511021230-8838
+LLM_DOCKER_IMAGE=urm.nvidia.com/sw-tensorrt-docker/tensorrt-llm:pytorch-25.10-py3-x86_64-ubuntu24.04-trt10.13.3.9-skip-tritondevel-202511110140-8447
+LLM_SBSA_DOCKER_IMAGE=urm.nvidia.com/sw-tensorrt-docker/tensorrt-llm:pytorch-25.10-py3-aarch64-ubuntu24.04-trt10.13.3.9-skip-tritondevel-202511110140-8447
+LLM_ROCKYLINUX8_PY310_DOCKER_IMAGE=urm.nvidia.com/sw-tensorrt-docker/tensorrt-llm:cuda-13.0.2-devel-rocky8-x86_64-rocky8-py310-trt10.13.3.9-skip-tritondevel-202511110140-8447
+LLM_ROCKYLINUX8_PY312_DOCKER_IMAGE=urm.nvidia.com/sw-tensorrt-docker/tensorrt-llm:cuda-13.0.2-devel-rocky8-x86_64-rocky8-py312-trt10.13.3.9-skip-tritondevel-202511110140-8447