ghcr.io/llm-d/llm-d-cuda:v0.8.0 - 镜像下载

ghcr.io/llm-d/llm-d-cuda:v0.8.0

linux/amd64 ghcr.io

这是一个用于大语言模型（LLM）相关任务的Docker容器镜像，基于CUDA技术支持GPU加速，适用于需要利用GPU资源运行大语言模型相关应用或服务的场景。

浏览次数

16.98GB

镜像大小

国内镜像

swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0

源镜像

ghcr.io/llm-d/llm-d-cuda:v0.8.0

镜像ID

sha256:f5924da0519222be33ea8683576ad2816d3a0c5519df9ab37b4143637cb2ee9e

镜像 TAG

v0.8.0

镜像大小

16.98GB

平台架构

linux/amd64

镜像源

ghcr.io

CMD

启动入口

/opt/nvidia/nvidia_entrypoint.sh

工作目录

/home/vllm

OS/平台

linux/amd64

镜像创建

2026-06-24T21:44:54.733497809Z

同步时间

2026-08-22 01:17

浏览量

4 次

贡献者

⚙️ 环境变量 53

Key	Value
`PATH=/opt/vllm/bin:/usr/local/cuda/bin:/usr/local/nvidia/bin:/opt/ucx/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin`	`0`
`container=oci`	`1`
`PKG_CMD=yum`	`2`
`NVARCH=x86_64`	`3`
NVIDIA_REQUIRE_CUDA=cuda>=13.0 brand=unknown,driver>=535,driver<536 brand=grid,driver>=535,driver<536 brand=tesla,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=vapps,driver>=535,driver<536 brand=vpc,driver>=535,driver<536 brand=vcs,driver>=535,driver<536 brand=vws,driver>=535,driver<536 brand=cloudgaming,driver>=535,driver<536 brand=unknown,driver>=550,driver<551 brand=grid,driver>=550,driver<551 brand=tesla,driver>=550,driver<551 brand=nvidia,driver>=550,driver<551 brand=quadro,driver>=550,driver<551 brand=quadrortx,driver>=550,driver<551 brand=nvidiartx,driver>=550,driver<551 brand=vapps,driver>=550,driver<551 brand=vpc,driver>=550,driver<551 brand=vcs,driver>=550,driver<551 brand=vws,driver>=550,driver<551 brand=cloudgaming,driver>=550,driver<551 brand=unknown,driver>=565,driver<566 brand=grid,driver>=565,driver<566 brand=tesla,driver>=565,driver<566 brand=nvidia,driver>=565,driver<566 brand=quadro,driver>=565,driver<566 brand=quadrortx,driver>=565,driver<566 brand=nvidiartx,driver>=565,driver<566 brand=vapps,driver>=565,driver<566 brand=vpc,driver>=565,driver<566 brand=vcs,driver>=565,driver<566 brand=vws,driver>=565,driver<566 brand=cloudgaming,driver>=565,driver<566 brand=unknown,driver>=570,driver<571 brand=grid,driver>=570,driver<571 brand=tesla,driver>=570,driver<571 brand=nvidia,driver>=570,driver<571 brand=quadro,driver>=570,driver<571 brand=quadrortx,driver>=570,driver<571 brand=nvidiartx,driver>=570,driver<571 brand=vapps,driver>=570,driver<571 brand=vpc,driver>=570,driver<571 brand=vcs,driver>=570,driver<571 brand=vws,driver>=570,driver<571 brand=cloudgaming,driver>=570,driver<571 brand=unknown,driver>=575,driver<576 brand=grid,driver>=575,driver<576 brand=tesla,driver>=575,driver<576 brand=nvidia,driver>=575,driver<576 brand=quadro,driver>=575,driver<576 brand=quadrortx,driver>=575,driver<576 brand=nvidiartx,driver>=575,driver<576 brand=vapps,driver>=575,driver<576 brand=vpc,driver>=575,driver<576 brand=vcs,driver>=575,driver<576 brand=vws,driver>=575,driver<576 brand=cloudgaming,driver>=575,driver<576	`4`
`NV_CUDA_CUDART_VERSION=13.0.96-1`	`5`
`CUDA_VERSION=13.0.2`	`6`
`LD_LIBRARY_PATH=/opt/nvshmem-v3.4.5-0/lib:/opt/nvshmem-v3.4.5-0/lib64:/opt/vllm/lib/python3.12/site-packages/nvidia/nvshmem/lib:/opt/vllm/lib64/python3.12/site-packages/torch/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/opt/amazon/efa/lib:/opt/amazon/efa/lib64:/opt/ucx/lib:/opt/ucx/lib64:/opt/uccl/lib:/usr/local/lib:/usr/local/lib64:/usr/lib/x86_64-linux-gnu:/usr/lib/aarch64-linux-gnu:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64`	`7`
`NVIDIA_VISIBLE_DEVICES=all`	`8`
`NVIDIA_DRIVER_CAPABILITIES=compute,utility`	`9`
`NV_CUDA_LIB_VERSION=13.0.2-1`	`10`
`NV_NVTX_VERSION=13.0.85-1`	`11`
`NV_LIBNPP_VERSION=13.0.1.2-1`	`12`
`NV_LIBNPP_PACKAGE=libnpp-13-0-13.0.1.2-1`	`13`
`NV_LIBCUBLAS_VERSION=13.1.0.3-1`	`14`
`NV_LIBNCCL_PACKAGE_NAME=libnccl`	`15`
`NV_LIBNCCL_PACKAGE_VERSION=2.28.3-1`	`16`
`NV_LIBNCCL_VERSION=2.28.3`	`17`
`NCCL_VERSION=2.28.3`	`18`
`NV_LIBNCCL_PACKAGE=libnccl-2.28.3-1+cuda13.0`	`19`
`NVIDIA_PRODUCT_NAME=CUDA`	`20`
`LANG=C.UTF-8`	`21`
`MAX_JOBS=3`	`22`
`LC_ALL=C.UTF-8`	`23`
`PYTHON_VERSION=3.12`	`24`
`UV_TORCH_BACKEND=cu130`	`25`
`UV_CONSTRAINT=/tmp/constraints.txt`	`26`
`VIRTUAL_ENV=/opt/vllm`	`27`
`NVSHMEM_DIR=/opt/nvshmem-v3.4.5-0`	`28`
`UCX_PREFIX=/opt/ucx`	`29`
`UCCL_PREFIX=/opt/uccl`	`30`
`EFA_PREFIX=/opt/amazon/efa`	`31`
`CUDA_HOME=/usr/local/cuda`	`32`
`CPATH=/usr/local/cuda/include:/usr/local/cuda/include/cccl:`	`33`
`TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX`	`34`
`FLASHINFER_VERSION=v0.6.12`	`35`
`CUDA_MAJOR=13`	`36`
`CUDA_MINOR=0`	`37`
`LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/opt/amazon/efa/lib:/opt/amazon/efa/lib64:/opt/ucx/lib:/opt/ucx/lib64:/opt/nvshmem-v3.4.5-0/lib:/opt/nvshmem-v3.4.5-0/lib64:/usr/local/lib:/usr/local/lib64:`	`38`
`LLM_D_MODELS_DIR=/var/lib/llm-d/models`	`39`
`HF_HOME=/var/lib/llm-d/.hf`	`40`
`HOME=/home/vllm`	`41`
`VLLM_USAGE_SOURCE=production-docker-image`	`42`
`VLLM_WORKER_MULTIPROC_METHOD=fork`	`43`
`OUTLINES_CACHE_DIR=/tmp/outlines`	`44`
`NUMBA_CACHE_DIR=/tmp/numba`	`45`
`TRITON_CACHE_DIR=/tmp/triton`	`46`
`TRITON_LIBCUDA_PATH=/usr/lib64`	`47`
`TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=15`	`48`
`TORCH_NCCL_DUMP_ON_TIMEOUT=0`	`49`
`VLLM_SKIP_P2P_CHECK=1`	`50`
`VLLM_CACHE_ROOT=/tmp/vllm`	`51`
`UCX_MEM_MMAP_HOOK_MODE=none`	`52`

🏷️ 镜像标签 29

Key	Value
`x86_64`	`architecture`
`2025-10-13T07:36:46Z`	`build-date`
`ubi9-container`	`com.redhat.component`
`https://www.redhat.com/en/about/red-hat-end-user-license-agreements#UBI`	`com.redhat.license_terms`
`cpe:/a:redhat:enterprise_linux:9::appstream`	`cpe`
`The Universal Base Image is designed and engineered to be the base layer for all of your containerized applications, middleware and utilities. This base image is freely redistributable, but Red Hat only supports Red Hat technologies through subscriptions for Red Hat products. This image is maintained by Red Hat and updated regularly.`	`description`
`public`	`distribution-scope`
`1.41.4`	`io.buildah.version`
`The Universal Base Image is designed and engineered to be the base layer for all of your containerized applications, middleware and utilities. This base image is freely redistributable, but Red Hat only supports Red Hat technologies through subscriptions for Red Hat products. This image is maintained by Red Hat and updated regularly.`	`io.k8s.description`
`Red Hat Universal Base Image 9`	`io.k8s.display-name`
	`io.openshift.expose-services`
`base rhel9`	`io.openshift.tags`
`NVIDIA CORPORATION <sw-cuda-installer@nvidia.com>`	`maintainer`
`ubi9/ubi`	`name`
`2026-06-24T21:21:44.196Z`	`org.opencontainers.image.created`
`Achieve state of the art inference performance with modern accelerators on Kubernetes`	`org.opencontainers.image.description`
`Apache-2.0`	`org.opencontainers.image.licenses`
`6afd09cc686309e3fd721cd9d88db8654c2c079b`	`org.opencontainers.image.revision`
`https://github.com/llm-d/llm-d`	`org.opencontainers.image.source`
`llm-d`	`org.opencontainers.image.title`
`https://github.com/llm-d/llm-d`	`org.opencontainers.image.url`
`v0.8.0`	`org.opencontainers.image.version`
`1760340943`	`release`
`Provides the latest release of Red Hat Universal Base Image 9.`	`summary`
`https://catalog.redhat.com/en/search?searchType=containers`	`url`
`60d587d1286655c5e777e63959aaae224123ea95`	`vcs-ref`
`git`	`vcs-type`
`Red Hat, Inc.`	`vendor`
`9.6`	`version`

🛡️ 镜像安全扫描

redhat 9.8 Trivy 2026-08-22 01:23 查看完整报告

1862

低危 LOW

3348

中危 MEDIUM

133

高危 HIGH

严重 CRITICAL

受影响目标 (5)

ghcr.io/llm-d/llm-d-cuda:v0.8.0 (redhat 9.8) redhat Python python-pkg opt/vllm-source/rust/Cargo.lock cargo usr/local/bin/uv rustbinary usr/local/bin/uvx rustbinary

Chatgpt|Claude一站式支持平台提供先进的AI镜像服务，助你在国内优雅的使用Chatgpt和Claude等AI服务

无权限下载？点我修复生成 Docker Run 命令生成 Kubernetes YAML

Docker拉取命令

docker pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0
docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0  ghcr.io/llm-d/llm-d-cuda:v0.8.0

Containerd拉取命令

ctr images pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0
ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0  ghcr.io/llm-d/llm-d-cuda:v0.8.0

Shell快速替换命令

sed -i 's#ghcr.io/llm-d/llm-d-cuda:v0.8.0#swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0#' deployment.yaml

Ansible快速分发-Docker

#ansible k8s -m shell -a 'docker pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0 && docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0  ghcr.io/llm-d/llm-d-cuda:v0.8.0'

Ansible快速分发-Containerd

#ansible k8s -m shell -a 'ctr images pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0 && ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0  ghcr.io/llm-d/llm-d-cuda:v0.8.0'

镜像构建历史


# 2026-06-25 05:44:54  0.00B 设置工作目录为/home/vllm
WORKDIR /home/vllm
                        
# 2026-06-25 05:44:54  0.00B 指定运行容器时使用的用户
USER 2000
                        
# 2026-06-25 05:44:54  0.00B 设置环境变量 PATH HOME VLLM_USAGE_SOURCE VLLM_WORKER_MULTIPROC_METHOD OUTLINES_CACHE_DIR NUMBA_CACHE_DIR TRITON_CACHE_DIR TRITON_LIBCUDA_PATH TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC TORCH_NCCL_DUMP_ON_TIMEOUT VLLM_SKIP_P2P_CHECK VLLM_CACHE_ROOT UCX_MEM_MMAP_HOOK_MODE
ENV PATH=/opt/vllm/bin:/usr/local/cuda/bin:/usr/local/nvidia/bin:/opt/ucx/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin HOME=/home/vllm VLLM_USAGE_SOURCE=production-docker-image VLLM_WORKER_MULTIPROC_METHOD=fork OUTLINES_CACHE_DIR=/tmp/outlines NUMBA_CACHE_DIR=/tmp/numba TRITON_CACHE_DIR=/tmp/triton TRITON_LIBCUDA_PATH=/usr/lib64 TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=15 TORCH_NCCL_DUMP_ON_TIMEOUT=0 VLLM_SKIP_P2P_CHECK=1 VLLM_CACHE_ROOT=/tmp/vllm UCX_MEM_MMAP_HOOK_MODE=none
                        
# 2026-06-25 05:44:54  21.00B 执行命令并创建新的镜像层
RUN |25 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true VLLM_REPO=https://github.com/neuralmagic/vllm.git VLLM_COMMIT_SHA=51f799c1a0c8a0476faf7e17eeb6a77983cdd778 VLLM_PRECOMPILED_WHEEL_COMMIT=0fc695fc6d1d82e9a5ac6835ac8e4e1c83703665 VLLM_PREBUILT=0 VLLM_USE_PRECOMPILED=1 SUPPRESS_PYTHON_OUTPUT= /bin/sh -c mkdir -p "$LLM_D_MODELS_DIR" "$HF_HOME" &&     chown -R root:0 /var/lib/llm-d &&     chmod -R g+rwX /var/lib/llm-d &&     find /var/lib/llm-d -type d -exec chmod g+s {} \; &&     ln -snf /var/lib/llm-d/models /models # buildkit
                        
# 2026-06-25 05:44:54  0.00B 设置环境变量 LLM_D_MODELS_DIR HF_HOME
ENV LLM_D_MODELS_DIR=/var/lib/llm-d/models HF_HOME=/var/lib/llm-d/.hf
                        
# 2026-06-25 05:44:54  0.00B 执行命令并创建新的镜像层
RUN |25 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true VLLM_REPO=https://github.com/neuralmagic/vllm.git VLLM_COMMIT_SHA=51f799c1a0c8a0476faf7e17eeb6a77983cdd778 VLLM_PRECOMPILED_WHEEL_COMMIT=0fc695fc6d1d82e9a5ac6835ac8e4e1c83703665 VLLM_PREBUILT=0 VLLM_USE_PRECOMPILED=1 SUPPRESS_PYTHON_OUTPUT= /bin/sh -c umask 002 &&     rm -rf /home/vllm &&     mkdir -p /home/vllm &&     chown vllm:root /home/vllm &&     chmod g+rwx /home/vllm # buildkit
                        
# 2026-06-25 05:44:54  9.34MB 执行命令并创建新的镜像层
RUN |25 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true VLLM_REPO=https://github.com/neuralmagic/vllm.git VLLM_COMMIT_SHA=51f799c1a0c8a0476faf7e17eeb6a77983cdd778 VLLM_PRECOMPILED_WHEEL_COMMIT=0fc695fc6d1d82e9a5ac6835ac8e4e1c83703665 VLLM_PREBUILT=0 VLLM_USE_PRECOMPILED=1 SUPPRESS_PYTHON_OUTPUT= /bin/sh -c bash -c ". /tmp/package-utils.sh &&     autoremove_packages \${TARGETOS} &&     cleanup_packages \${TARGETOS} &&     rm /tmp/package-utils.sh" # buildkit
                        
# 2026-06-25 05:44:39  6.92KB 复制新文件或目录到容器中
COPY docker/scripts/cuda/common/package-utils.sh /tmp/package-utils.sh # buildkit
                        
# 2026-06-25 05:44:39  590.85KB 执行命令并创建新的镜像层
RUN |25 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true VLLM_REPO=https://github.com/neuralmagic/vllm.git VLLM_COMMIT_SHA=51f799c1a0c8a0476faf7e17eeb6a77983cdd778 VLLM_PRECOMPILED_WHEEL_COMMIT=0fc695fc6d1d82e9a5ac6835ac8e4e1c83703665 VLLM_PREBUILT=0 VLLM_USE_PRECOMPILED=1 SUPPRESS_PYTHON_OUTPUT= /bin/sh -c uv pip install -r /workspace/runtime-otel-package-requirements.txt # buildkit
                        
# 2026-06-25 05:44:39  180.00B 复制新文件或目录到容器中
COPY docker/packages/common/runtime-otel-package-requirements.txt /workspace/runtime-otel-package-requirements.txt # buildkit
                        
# 2026-06-25 05:44:39  33.16KB 执行命令并创建新的镜像层
RUN |25 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true VLLM_REPO=https://github.com/neuralmagic/vllm.git VLLM_COMMIT_SHA=51f799c1a0c8a0476faf7e17eeb6a77983cdd778 VLLM_PRECOMPILED_WHEEL_COMMIT=0fc695fc6d1d82e9a5ac6835ac8e4e1c83703665 VLLM_PREBUILT=0 VLLM_USE_PRECOMPILED=1 SUPPRESS_PYTHON_OUTPUT= /bin/sh -c if [ "${NVSHMEM_BUILD_FROM_SOURCE}" = "true" ]; then         echo "/opt/nvshmem-${NVSHMEM_VERSION}/lib" > /etc/ld.so.conf.d/nvidia-nvshmem.conf &&         echo "/opt/nvshmem-${NVSHMEM_VERSION}/lib64" >> /etc/ld.so.conf.d/nvidia-nvshmem.conf;     else         echo "/opt/vllm/lib/python${PYTHON_VERSION}/site-packages/nvidia/nvshmem/lib"             > /etc/ld.so.conf.d/nvidia-nvshmem.conf;     fi &&     echo "/opt/vllm/lib/python${PYTHON_VERSION}/site-packages/nvidia/nccl/lib"         >> /etc/ld.so.conf.d/nvidia-nvshmem.conf &&     ldconfig # buildkit
                        
# 2026-06-25 05:44:38  11.91GB 执行命令并创建新的镜像层
RUN |25 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true VLLM_REPO=https://github.com/neuralmagic/vllm.git VLLM_COMMIT_SHA=51f799c1a0c8a0476faf7e17eeb6a77983cdd778 VLLM_PRECOMPILED_WHEEL_COMMIT=0fc695fc6d1d82e9a5ac6835ac8e4e1c83703665 VLLM_PREBUILT=0 VLLM_USE_PRECOMPILED=1 SUPPRESS_PYTHON_OUTPUT= /bin/sh -c bash -c "chmod +x /tmp/install-vllm.sh &&     export PATH=/usr/local/bin:\${PATH} &&     export NVSHMEM_BUILD_FROM_SOURCE=${NVSHMEM_BUILD_FROM_SOURCE} &&     USE_SCCACHE=${USE_SCCACHE} TARGETPLATFORM=${TARGETPLATFORM} source /usr/local/bin/setup-sccache &&     /tmp/install-vllm.sh &&     rm /tmp/install-vllm.sh &&     if [ \"\${USE_SCCACHE}\" = \"true\" ]; then sccache --show-stats || true; fi" # buildkit
                        
# 2026-06-25 05:42:30  7.56KB 复制新文件或目录到容器中
COPY docker/scripts/cuda/runtime/install-vllm.sh /tmp/install-vllm.sh # buildkit
                        
# 2026-06-25 05:42:30  1.47KB 复制新文件或目录到容器中
COPY scripts/warn-vllm-precompiled.sh /opt/ # buildkit
                        
# 2026-06-25 05:42:30  0.00B 定义构建参数
ARG SUPPRESS_PYTHON_OUTPUT
                        
# 2026-06-25 05:42:30  0.00B 定义构建参数
ARG VLLM_USE_PRECOMPILED=1
                        
# 2026-06-25 05:42:30  0.00B 定义构建参数
ARG VLLM_PREBUILT=0
                        
# 2026-06-25 05:42:30  0.00B 定义构建参数
ARG VLLM_PRECOMPILED_WHEEL_COMMIT=0fc695fc6d1d82e9a5ac6835ac8e4e1c83703665
                        
# 2026-06-25 05:42:30  0.00B 定义构建参数
ARG VLLM_COMMIT_SHA=51f799c1a0c8a0476faf7e17eeb6a77983cdd778
                        
# 2026-06-25 05:42:30  0.00B 定义构建参数
ARG VLLM_REPO=https://github.com/neuralmagic/vllm.git
                        
# 2026-06-25 05:42:30  0.00B 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c mkdir -p /opt/vllm-source &&     chown -R 2000:0 /opt/vllm-source &&     chmod -R g+rwX /opt/vllm-source &&     find /opt/vllm-source -type d -exec chmod g+s {} \; &&     setfacl -R -m g:0:rwX -m d:g:0:rwX /opt/vllm-source || true # buildkit
                        
# 2026-06-25 05:42:30  587.28KB 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c useradd --uid 2000 --gid 0 -m vllm &&     touch /home/vllm/.bashrc # buildkit
                        
# 2026-06-25 05:42:30  43.44MB 复制新文件或目录到容器中
COPY /wheels/*.whl /tmp/wheels/ # buildkit
                        
# 2026-06-25 05:42:30  2.69KB 复制新文件或目录到容器中
COPY /usr/local/bin/setup-sccache /usr/local/bin/setup-sccache # buildkit
                        
# 2026-06-25 05:42:30  21.80MB 复制新文件或目录到容器中
COPY /usr/local/bin/sccache /usr/local/bin/sccache # buildkit
                        
# 2026-06-25 05:42:30  62.19MB 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c export UV_INSTALL_DIR=/usr/local/bin &&     curl -LsSf https://astral.sh/uv/install.sh | sh &&     uv venv /opt/vllm --python ${PYTHON_VERSION} &&     rm -f /usr/bin/python3 /usr/bin/python3-config /usr/bin/pip &&     ln -s /opt/vllm/bin/python3 /usr/bin/python3 &&     ln -s /opt/vllm/bin/python3-config /usr/bin/python3-config &&     ln -s /opt/vllm/bin/pip /usr/bin/pip &&     uv pip install --no-cache -U wheel # buildkit
                        
# 2026-06-25 05:42:29  32.72KB 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c echo "/usr/local/lib" > /etc/ld.so.conf.d/local.conf &&     echo "/usr/local/lib64" >> /etc/ld.so.conf.d/local.conf &&     ldconfig # buildkit
                        
# 2026-06-25 05:42:29  0.00B 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c if [ -d /opt/nixl/include ]; then         cp -a /opt/nixl/include/nixl* /usr/local/include/ || true;     fi;     if [ -d /opt/nixl/lib ]; then         if [ "${TARGETOS:-rhel}" = "ubuntu" ]; then             if [ "${TARGETPLATFORM:-linux/amd64}" = "linux/arm64" ]; then                 cp -a /opt/nixl/lib/aarch64-linux-gnu /usr/lib/aarch64-linux-gnu/ || true;             else                 cp -a /opt/nixl/lib/x86_64-linux-gnu /usr/lib/x86_64-linux-gnu/ || true;             fi;         else             cp -a /opt/nixl/lib64/ /usr/lib64/ || true;         fi;     fi;     rm -rf /opt/nixl # buildkit
                        
# 2026-06-25 05:42:29  0.00B 复制新文件或目录到容器中
COPY /opt/nixl /opt/nixl # buildkit
                        
# 2026-06-25 05:42:29  91.20MB 复制新文件或目录到容器中
COPY /opt/nvshmem-v3.4.5-0/ /opt/nvshmem-v3.4.5-0/ # buildkit
                        
# 2026-06-25 05:42:29  0.00B 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c if [ -d /tmp/gdrcopy_libs ]; then         if [ "${TARGETOS:-rhel}" = "ubuntu" ]; then             if [ "${TARGETPLATFORM:-linux/amd64}" = "linux/arm64" ]; then                 cp -a /tmp/gdrcopy_libs/* /usr/lib/aarch64-linux-gnu/ || true;             else                 cp -a /tmp/gdrcopy_libs/* /usr/lib/x86_64-linux-gnu/ || true;             fi;         else             cp -a /tmp/gdrcopy_libs/* /usr/lib64/ || true;         fi;         rm -rf /tmp/gdrcopy_libs;     fi # buildkit
                        
# 2026-06-25 05:42:28  1.14MB 复制新文件或目录到容器中
COPY /opt/uccl /opt/uccl # buildkit
                        
# 2026-06-25 05:42:28  4.91MB 复制新文件或目录到容器中
COPY /opt/ucx /opt/ucx # buildkit
                        
# 2026-06-25 05:42:28  31.34KB 复制新文件或目录到容器中
COPY /tmp/gdrcopy_libs /tmp/ # buildkit
                        
# 2026-06-25 05:42:28  0.00B 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c if [ "${TARGETOS:-rhel}" = "ubuntu" ]; then         if [ "${TARGETPLATFORM:-linux/amd64}" = "linux/arm64" ]; then             LIBDIR="/usr/lib/aarch64-linux-gnu";         else             LIBDIR="/usr/lib/x86_64-linux-gnu";         fi;     else         LIBDIR="/usr/lib64";     fi &&     mkdir -p "${LIBDIR}" # buildkit
                        
# 2026-06-25 05:42:28  94.03KB 复制新文件或目录到容器中
COPY /usr/local/lib/libgdrapi.so* /usr/local/lib/ # buildkit
                        
# 2026-06-25 05:42:28  0.00B 复制新文件或目录到容器中
COPY /tmp/efa_libs/ /usr/lib64/ # buildkit
                        
# 2026-06-25 05:42:28  0.00B 复制新文件或目录到容器中
COPY /opt/amazon /opt/amazon/ # buildkit
                        
# 2026-06-25 05:42:28  13.34MB 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c if [ "${INSTALL_OFFLOADING_CONNECTOR}" = "true" ]; then         /tmp/install-offloading-connector.sh;     else         echo "Skipping offloading connector installation (INSTALL_OFFLOADING_CONNECTOR=${INSTALL_OFFLOADING_CONNECTOR})";     fi &&     rm -f /tmp/install-offloading-connector.sh # buildkit
                        
# 2026-06-25 05:42:25  3.38KB 复制新文件或目录到容器中
COPY docker/scripts/cuda/runtime/install-offloading-connector.sh /tmp/install-offloading-connector.sh # buildkit
                        
# 2026-06-25 05:42:25  2.07GB 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c /tmp/install-runtime-packages.sh &&     rm -f /tmp/install-runtime-packages.sh /tmp/package-utils.sh &&     rm -rf /tmp/packages # buildkit
                        
# 2026-06-25 05:41:06  2.36KB 复制新文件或目录到容器中
COPY docker/scripts/cuda/runtime/install-runtime-packages.sh /tmp/install-runtime-packages.sh # buildkit
                        
# 2026-06-25 05:41:06  9.96MB 复制新文件或目录到容器中
COPY docker/packages /tmp/packages # buildkit
                        
# 2026-06-25 05:41:06  6.92KB 复制新文件或目录到容器中
COPY docker/scripts/cuda/common/package-utils.sh /tmp/package-utils.sh # buildkit
                        
# 2026-06-25 05:41:06  0.00B 执行命令并创建新的镜像层
RUN |19 CACHE_BUSTER=v0.8.0-28130361889 CUDA_MAJOR=13 CUDA_MINOR=0 CUDA_PATCH=2 TARGETOS=rhel TARGETPLATFORM=linux/amd64 FINAL_BASE_IMAGE_SUFFIX=ubi9 PYTHON_VERSION=3.12 NVSHMEM_VERSION=v3.4.5-0 NVSHMEM_BUILD_FROM_SOURCE=true BUILD_NIXL_FROM_SOURCE=false FLASHINFER_VERSION=v0.6.12 TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX USE_SCCACHE=true MAX_JOBS=3 ENABLE_EFA=false EFA_INSTALLER_VERSION=1.46.0 LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23 INSTALL_OFFLOADING_CONNECTOR=true /bin/sh -c EFA_MODE="runtime" /tmp/install-efa.sh &&     rm -f /tmp/install-efa.sh /tmp/package-utils.sh &&     rm -rf /tmp/packages # buildkit
                        
# 2026-06-25 05:41:06  3.86KB 复制新文件或目录到容器中
COPY docker/scripts/cuda/common/install-efa.sh /tmp/install-efa.sh # buildkit
                        
# 2026-06-25 05:41:06  9.96MB 复制新文件或目录到容器中
COPY docker/packages /tmp/packages # buildkit
                        
# 2026-06-25 05:41:06  6.92KB 复制新文件或目录到容器中
COPY docker/scripts/cuda/common/package-utils.sh /tmp/package-utils.sh # buildkit
                        
# 2026-06-25 05:41:06  0.00B 设置环境变量 LD_LIBRARY_PATH PATH CPATH TORCH_CUDA_ARCH_LIST FLASHINFER_VERSION CUDA_MAJOR CUDA_MINOR LIBRARY_PATH
ENV LD_LIBRARY_PATH=/opt/nvshmem-v3.4.5-0/lib:/opt/nvshmem-v3.4.5-0/lib64:/opt/vllm/lib/python3.12/site-packages/nvidia/nvshmem/lib:/opt/vllm/lib64/python3.12/site-packages/torch/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/opt/amazon/efa/lib:/opt/amazon/efa/lib64:/opt/ucx/lib:/opt/ucx/lib64:/opt/uccl/lib:/usr/local/lib:/usr/local/lib64:/usr/lib/x86_64-linux-gnu:/usr/lib/aarch64-linux-gnu:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64 PATH=/opt/ucx/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin CPATH=/usr/local/cuda/include:/usr/local/cuda/include/cccl: TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX FLASHINFER_VERSION=v0.6.12 CUDA_MAJOR=13 CUDA_MINOR=0 LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/opt/amazon/efa/lib:/opt/amazon/efa/lib64:/opt/ucx/lib:/opt/ucx/lib64:/opt/nvshmem-v3.4.5-0/lib:/opt/nvshmem-v3.4.5-0/lib64:/usr/local/lib:/usr/local/lib64:
                        
# 2026-06-25 05:41:06  0.00B 设置环境变量 LANG MAX_JOBS LC_ALL PYTHON_VERSION UV_TORCH_BACKEND UV_CONSTRAINT VIRTUAL_ENV NVSHMEM_DIR UCX_PREFIX UCCL_PREFIX EFA_PREFIX CUDA_HOME
ENV LANG=C.UTF-8 MAX_JOBS=3 LC_ALL=C.UTF-8 PYTHON_VERSION=3.12 UV_TORCH_BACKEND=cu130 UV_CONSTRAINT=/tmp/constraints.txt VIRTUAL_ENV=/opt/vllm NVSHMEM_DIR=/opt/nvshmem-v3.4.5-0 UCX_PREFIX=/opt/ucx UCCL_PREFIX=/opt/uccl EFA_PREFIX=/opt/amazon/efa CUDA_HOME=/usr/local/cuda
                        
# 2026-06-25 05:41:06  60.00B 复制新文件或目录到容器中
COPY docker/constraints.txt /tmp/constraints.txt # buildkit
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG INSTALL_OFFLOADING_CONNECTOR=true
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG LLM_D_OFFLOADING_CONNECTOR_VERSION=0.23
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG EFA_INSTALLER_VERSION=1.46.0
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG ENABLE_EFA=false
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG MAX_JOBS=3
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG USE_SCCACHE=true
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG FLASHINFER_VERSION=v0.6.12
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG BUILD_NIXL_FROM_SOURCE=false
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG NVSHMEM_BUILD_FROM_SOURCE=true
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG NVSHMEM_VERSION=v3.4.5-0
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG PYTHON_VERSION=3.12
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG FINAL_BASE_IMAGE_SUFFIX=ubi9
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG TARGETPLATFORM=linux/amd64
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG TARGETOS=rhel
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG CUDA_PATCH=2
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG CUDA_MINOR=0
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG CUDA_MAJOR=13
                        
# 2026-06-25 05:41:06  19.00B 执行命令并创建新的镜像层
RUN |1 CACHE_BUSTER=v0.8.0-28130361889 /bin/sh -c if [ -n "${CACHE_BUSTER}" ]; then         echo "$CACHE_BUSTER" > /tmp/builder-buster;     fi; # buildkit
                        
# 2026-06-25 05:41:06  0.00B 定义构建参数
ARG CACHE_BUSTER=v0.8.0-28130361889
                        
# 2025-11-05 07:53:21  0.00B 配置容器启动时运行的命令
ENTRYPOINT ["/opt/nvidia/nvidia_entrypoint.sh"]
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NVIDIA_PRODUCT_NAME
ENV NVIDIA_PRODUCT_NAME=CUDA
                        
# 2025-11-05 07:53:21  2.53KB 复制新文件或目录到容器中
COPY nvidia_entrypoint.sh /opt/nvidia/ # buildkit
                        
# 2025-11-05 07:53:21  3.06KB 复制新文件或目录到容器中
COPY entrypoint.d/ /opt/nvidia/entrypoint.d/ # buildkit
                        
# 2025-11-05 07:53:21  2.14GB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c ${PKG_CMD} install -y     cuda-libraries-13-0-${NV_CUDA_LIB_VERSION}     cuda-nvtx-13-0-${NV_NVTX_VERSION}     ${NV_LIBNPP_PACKAGE}     libcublas-13-0-${NV_LIBCUBLAS_VERSION}     ${NV_LIBNCCL_PACKAGE}     && ${PKG_CMD} clean all     && rm -rf /var/cache/yum/* # buildkit
                        
# 2025-11-05 07:53:21  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <sw-cuda-installer@nvidia.com>
                        
# 2025-11-05 07:53:21  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NV_LIBNCCL_PACKAGE
ENV NV_LIBNCCL_PACKAGE=libnccl-2.28.3-1+cuda13.0
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NCCL_VERSION
ENV NCCL_VERSION=2.28.3
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NV_LIBNCCL_VERSION
ENV NV_LIBNCCL_VERSION=2.28.3
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NV_LIBNCCL_PACKAGE_VERSION
ENV NV_LIBNCCL_PACKAGE_VERSION=2.28.3-1
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NV_LIBNCCL_PACKAGE_NAME
ENV NV_LIBNCCL_PACKAGE_NAME=libnccl
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NV_LIBCUBLAS_VERSION
ENV NV_LIBCUBLAS_VERSION=13.1.0.3-1
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NV_LIBNPP_PACKAGE
ENV NV_LIBNPP_PACKAGE=libnpp-13-0-13.0.1.2-1
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NV_LIBNPP_VERSION
ENV NV_LIBNPP_VERSION=13.0.1.2-1
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NV_NVTX_VERSION
ENV NV_NVTX_VERSION=13.0.85-1
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 NV_CUDA_LIB_VERSION
ENV NV_CUDA_LIB_VERSION=13.0.2-1
                        
# 2025-11-05 07:53:21  0.00B 设置环境变量 PKG_CMD
ENV PKG_CMD=yum
                        
# 2025-11-05 07:49:52  0.00B 设置环境变量 NVIDIA_DRIVER_CAPABILITIES
ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility
                        
# 2025-11-05 07:49:52  0.00B 设置环境变量 NVIDIA_VISIBLE_DEVICES
ENV NVIDIA_VISIBLE_DEVICES=all
                        
# 2025-11-05 07:49:52  17.29KB 复制新文件或目录到容器中
COPY NGC-DL-CONTAINER-LICENSE / # buildkit
                        
# 2025-11-05 07:49:51  0.00B 设置环境变量 LD_LIBRARY_PATH
ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64
                        
# 2025-11-05 07:49:51  0.00B 设置环境变量 PATH
ENV PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
                        
# 2025-11-05 07:49:51  22.00B 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c echo "/usr/local/cuda/lib64" >> /etc/ld.so.conf.d/nvidia.conf # buildkit
                        
# 2025-11-05 07:49:50  375.20MB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c ${PKG_CMD} upgrade -y && ${PKG_CMD} install -y     cuda-cudart-13-0-${NV_CUDA_CUDART_VERSION}     cuda-compat-13-0     && ${PKG_CMD} clean all     && rm -rf /var/cache/yum/* # buildkit
                        
# 2025-11-05 07:49:40  0.00B 设置环境变量 CUDA_VERSION
ENV CUDA_VERSION=13.0.2
                        
# 2025-11-05 07:49:40  1.61KB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=amd64 /bin/sh -c NVIDIA_GPGKEY_SUM=d0664fbbdb8c32356d45de36c5984617217b2d0bef41b93ccecd326ba3b80c87 &&     curl -fsSL https://developer.download.nvidia.com/compute/cuda/repos/rhel9/${NVARCH}/D42D0685.pub | sed '/^Version/d' > /etc/pki/rpm-gpg/RPM-GPG-KEY-NVIDIA &&     echo "$NVIDIA_GPGKEY_SUM  /etc/pki/rpm-gpg/RPM-GPG-KEY-NVIDIA" | sha256sum -c --strict - # buildkit
                        
# 2025-11-05 07:49:39  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <sw-cuda-installer@nvidia.com>
                        
# 2025-11-05 07:49:39  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2025-11-05 07:49:39  165.00B 复制新文件或目录到容器中
COPY cuda.repo-x86_64 /etc/yum.repos.d/cuda.repo # buildkit
                        
# 2025-11-05 07:49:39  0.00B 设置环境变量 NV_CUDA_CUDART_VERSION
ENV NV_CUDA_CUDART_VERSION=13.0.96-1
                        
# 2025-11-05 07:49:39  0.00B 设置环境变量 NVIDIA_REQUIRE_CUDA brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand brand
ENV NVIDIA_REQUIRE_CUDA=cuda>=13.0 brand=unknown,driver>=535,driver<536 brand=grid,driver>=535,driver<536 brand=tesla,driver>=535,driver<536 brand=nvidia,driver>=535,driver<536 brand=quadro,driver>=535,driver<536 brand=quadrortx,driver>=535,driver<536 brand=nvidiartx,driver>=535,driver<536 brand=vapps,driver>=535,driver<536 brand=vpc,driver>=535,driver<536 brand=vcs,driver>=535,driver<536 brand=vws,driver>=535,driver<536 brand=cloudgaming,driver>=535,driver<536 brand=unknown,driver>=550,driver<551 brand=grid,driver>=550,driver<551 brand=tesla,driver>=550,driver<551 brand=nvidia,driver>=550,driver<551 brand=quadro,driver>=550,driver<551 brand=quadrortx,driver>=550,driver<551 brand=nvidiartx,driver>=550,driver<551 brand=vapps,driver>=550,driver<551 brand=vpc,driver>=550,driver<551 brand=vcs,driver>=550,driver<551 brand=vws,driver>=550,driver<551 brand=cloudgaming,driver>=550,driver<551 brand=unknown,driver>=565,driver<566 brand=grid,driver>=565,driver<566 brand=tesla,driver>=565,driver<566 brand=nvidia,driver>=565,driver<566 brand=quadro,driver>=565,driver<566 brand=quadrortx,driver>=565,driver<566 brand=nvidiartx,driver>=565,driver<566 brand=vapps,driver>=565,driver<566 brand=vpc,driver>=565,driver<566 brand=vcs,driver>=565,driver<566 brand=vws,driver>=565,driver<566 brand=cloudgaming,driver>=565,driver<566 brand=unknown,driver>=570,driver<571 brand=grid,driver>=570,driver<571 brand=tesla,driver>=570,driver<571 brand=nvidia,driver>=570,driver<571 brand=quadro,driver>=570,driver<571 brand=quadrortx,driver>=570,driver<571 brand=nvidiartx,driver>=570,driver<571 brand=vapps,driver>=570,driver<571 brand=vpc,driver>=570,driver<571 brand=vcs,driver>=570,driver<571 brand=vws,driver>=570,driver<571 brand=cloudgaming,driver>=570,driver<571 brand=unknown,driver>=575,driver<576 brand=grid,driver>=575,driver<576 brand=tesla,driver>=575,driver<576 brand=nvidia,driver>=575,driver<576 brand=quadro,driver>=575,driver<576 brand=quadrortx,driver>=575,driver<576 brand=nvidiartx,driver>=575,driver<576 brand=vapps,driver>=575,driver<576 brand=vpc,driver>=575,driver<576 brand=vcs,driver>=575,driver<576 brand=vws,driver>=575,driver<576 brand=cloudgaming,driver>=575,driver<576
                        
# 2025-11-05 07:49:39  0.00B 设置环境变量 NVARCH
ENV NVARCH=x86_64
                        
# 2025-11-05 07:49:39  0.00B 设置环境变量 PKG_CMD
ENV PKG_CMD=yum
                        
# 2025-10-13 15:37:08  209.62MB 
/bin/sh -c #(nop) LABEL "architecture"="x86_64" "vcs-type"="git" "vcs-ref"="60d587d1286655c5e777e63959aaae224123ea95" "org.opencontainers.image.revision"="60d587d1286655c5e777e63959aaae224123ea95" "build-date"="2025-10-13T07:36:46Z" "release"="1760340943"org.opencontainers.image.revision=60d587d1286655c5e777e63959aaae224123ea95
                        
# 2025-10-13 15:37:08  0.00B 
/bin/sh -c #(nop) COPY file:a782d4cd79a0c61643bc1cbf3cd5798a42f1495fe330dde8598373d62cac62dd in /root/buildinfo/labels.json      
                        
# 2025-10-13 15:37:08  0.00B 
/bin/sh -c #(nop) COPY file:93583a9ebbaeff1e36b48820b647eea1eef523f6627dacfb0b21af79f5a41b35 in /root/buildinfo/content_manifests/content-sets.json      
                        
# 2025-10-13 15:37:07  0.00B 
/bin/sh -c #(nop) COPY file:93583a9ebbaeff1e36b48820b647eea1eef523f6627dacfb0b21af79f5a41b35 in /usr/share/buildinfo/content-sets.json      
                        
# 2025-10-13 15:37:07  0.00B 
/bin/sh -c #(nop) CMD ["/bin/bash"]
                        
# 2025-10-13 15:37:07  0.00B 
/bin/sh -c #(nop) COPY file:1376702515d596f414e3aa494e0daa6d408a6d2475c4aeca96bf9392f5287f69 in /etc/yum.repos.d/.      
                        
# 2025-10-13 15:37:07  0.00B 
/bin/sh -c #(nop) COPY dir:91d00d24ce3ba29551746ee5faca2e3a563eef65ecd488882ee5f2d4f984589d in /      
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) ENV container oci
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL io.openshift.tags="base rhel9"
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL io.openshift.expose-services=""
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL io.k8s.display-name="Red Hat Universal Base Image 9"
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL io.k8s.description="The Universal Base Image is designed and engineered to be the base layer for all of your containerized applications, middleware and utilities. This base image is freely redistributable, but Red Hat only supports Red Hat technologies through subscriptions for Red Hat products. This image is maintained by Red Hat and updated regularly."
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL description="The Universal Base Image is designed and engineered to be the base layer for all of your containerized applications, middleware and utilities. This base image is freely redistributable, but Red Hat only supports Red Hat technologies through subscriptions for Red Hat products. This image is maintained by Red Hat and updated regularly."
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL summary="Provides the latest release of Red Hat Universal Base Image 9."
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL com.redhat.license_terms="https://www.redhat.com/en/about/red-hat-end-user-license-agreements#UBI"
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL com.redhat.component="ubi9-container"       name="ubi9/ubi"       version="9.6"       cpe="cpe:/a:redhat:enterprise_linux:9::appstream"       distribution-scope="public"
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL url="https://catalog.redhat.com/en/search?searchType=containers"
                        
# 2025-10-13 15:37:06  0.00B 
/bin/sh -c #(nop) LABEL maintainer="Red Hat, Inc."       vendor="Red Hat, Inc."

镜像信息

{
    "Id": "sha256:f5924da0519222be33ea8683576ad2816d3a0c5519df9ab37b4143637cb2ee9e",
    "RepoTags": [
        "ghcr.io/llm-d/llm-d-cuda:v0.8.0",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda:v0.8.0"
    ],
    "RepoDigests": [
        "ghcr.io/llm-d/llm-d-cuda@sha256:154469797de1a1838442660d6d4c6c67240aae77ca15fb37d5ade4c732d2d976",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/ghcr.io/llm-d/llm-d-cuda@sha256:fad9a2f712d3303ebce6616ece9b352ed2e13cbc8f07ea883af191c43b3495b2"
    ],
    "Parent": "",
    "Comment": "buildkit.dockerfile.v0",
    "Created": "2026-06-24T21:44:54.733497809Z",
    "Container": "",
    "ContainerConfig": null,
    "DockerVersion": "",
    "Author": "",
    "Config": {
        "Hostname": "",
        "Domainname": "",
        "User": "2000",
        "AttachStdin": false,
        "AttachStdout": false,
        "AttachStderr": false,
        "Tty": false,
        "OpenStdin": false,
        "StdinOnce": false,
        "Env": [
            "PATH=/opt/vllm/bin:/usr/local/cuda/bin:/usr/local/nvidia/bin:/opt/ucx/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin",
            "container=oci",
            "PKG_CMD=yum",
            "NVARCH=x86_64",
            "NVIDIA_REQUIRE_CUDA=cuda\u003e=13.0 brand=unknown,driver\u003e=535,driver\u003c536 brand=grid,driver\u003e=535,driver\u003c536 brand=tesla,driver\u003e=535,driver\u003c536 brand=nvidia,driver\u003e=535,driver\u003c536 brand=quadro,driver\u003e=535,driver\u003c536 brand=quadrortx,driver\u003e=535,driver\u003c536 brand=nvidiartx,driver\u003e=535,driver\u003c536 brand=vapps,driver\u003e=535,driver\u003c536 brand=vpc,driver\u003e=535,driver\u003c536 brand=vcs,driver\u003e=535,driver\u003c536 brand=vws,driver\u003e=535,driver\u003c536 brand=cloudgaming,driver\u003e=535,driver\u003c536 brand=unknown,driver\u003e=550,driver\u003c551 brand=grid,driver\u003e=550,driver\u003c551 brand=tesla,driver\u003e=550,driver\u003c551 brand=nvidia,driver\u003e=550,driver\u003c551 brand=quadro,driver\u003e=550,driver\u003c551 brand=quadrortx,driver\u003e=550,driver\u003c551 brand=nvidiartx,driver\u003e=550,driver\u003c551 brand=vapps,driver\u003e=550,driver\u003c551 brand=vpc,driver\u003e=550,driver\u003c551 brand=vcs,driver\u003e=550,driver\u003c551 brand=vws,driver\u003e=550,driver\u003c551 brand=cloudgaming,driver\u003e=550,driver\u003c551 brand=unknown,driver\u003e=565,driver\u003c566 brand=grid,driver\u003e=565,driver\u003c566 brand=tesla,driver\u003e=565,driver\u003c566 brand=nvidia,driver\u003e=565,driver\u003c566 brand=quadro,driver\u003e=565,driver\u003c566 brand=quadrortx,driver\u003e=565,driver\u003c566 brand=nvidiartx,driver\u003e=565,driver\u003c566 brand=vapps,driver\u003e=565,driver\u003c566 brand=vpc,driver\u003e=565,driver\u003c566 brand=vcs,driver\u003e=565,driver\u003c566 brand=vws,driver\u003e=565,driver\u003c566 brand=cloudgaming,driver\u003e=565,driver\u003c566 brand=unknown,driver\u003e=570,driver\u003c571 brand=grid,driver\u003e=570,driver\u003c571 brand=tesla,driver\u003e=570,driver\u003c571 brand=nvidia,driver\u003e=570,driver\u003c571 brand=quadro,driver\u003e=570,driver\u003c571 brand=quadrortx,driver\u003e=570,driver\u003c571 brand=nvidiartx,driver\u003e=570,driver\u003c571 brand=vapps,driver\u003e=570,driver\u003c571 brand=vpc,driver\u003e=570,driver\u003c571 brand=vcs,driver\u003e=570,driver\u003c571 brand=vws,driver\u003e=570,driver\u003c571 brand=cloudgaming,driver\u003e=570,driver\u003c571 brand=unknown,driver\u003e=575,driver\u003c576 brand=grid,driver\u003e=575,driver\u003c576 brand=tesla,driver\u003e=575,driver\u003c576 brand=nvidia,driver\u003e=575,driver\u003c576 brand=quadro,driver\u003e=575,driver\u003c576 brand=quadrortx,driver\u003e=575,driver\u003c576 brand=nvidiartx,driver\u003e=575,driver\u003c576 brand=vapps,driver\u003e=575,driver\u003c576 brand=vpc,driver\u003e=575,driver\u003c576 brand=vcs,driver\u003e=575,driver\u003c576 brand=vws,driver\u003e=575,driver\u003c576 brand=cloudgaming,driver\u003e=575,driver\u003c576",
            "NV_CUDA_CUDART_VERSION=13.0.96-1",
            "CUDA_VERSION=13.0.2",
            "LD_LIBRARY_PATH=/opt/nvshmem-v3.4.5-0/lib:/opt/nvshmem-v3.4.5-0/lib64:/opt/vllm/lib/python3.12/site-packages/nvidia/nvshmem/lib:/opt/vllm/lib64/python3.12/site-packages/torch/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/opt/amazon/efa/lib:/opt/amazon/efa/lib64:/opt/ucx/lib:/opt/ucx/lib64:/opt/uccl/lib:/usr/local/lib:/usr/local/lib64:/usr/lib/x86_64-linux-gnu:/usr/lib/aarch64-linux-gnu:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64",
            "NVIDIA_VISIBLE_DEVICES=all",
            "NVIDIA_DRIVER_CAPABILITIES=compute,utility",
            "NV_CUDA_LIB_VERSION=13.0.2-1",
            "NV_NVTX_VERSION=13.0.85-1",
            "NV_LIBNPP_VERSION=13.0.1.2-1",
            "NV_LIBNPP_PACKAGE=libnpp-13-0-13.0.1.2-1",
            "NV_LIBCUBLAS_VERSION=13.1.0.3-1",
            "NV_LIBNCCL_PACKAGE_NAME=libnccl",
            "NV_LIBNCCL_PACKAGE_VERSION=2.28.3-1",
            "NV_LIBNCCL_VERSION=2.28.3",
            "NCCL_VERSION=2.28.3",
            "NV_LIBNCCL_PACKAGE=libnccl-2.28.3-1+cuda13.0",
            "NVIDIA_PRODUCT_NAME=CUDA",
            "LANG=C.UTF-8",
            "MAX_JOBS=3",
            "LC_ALL=C.UTF-8",
            "PYTHON_VERSION=3.12",
            "UV_TORCH_BACKEND=cu130",
            "UV_CONSTRAINT=/tmp/constraints.txt",
            "VIRTUAL_ENV=/opt/vllm",
            "NVSHMEM_DIR=/opt/nvshmem-v3.4.5-0",
            "UCX_PREFIX=/opt/ucx",
            "UCCL_PREFIX=/opt/uccl",
            "EFA_PREFIX=/opt/amazon/efa",
            "CUDA_HOME=/usr/local/cuda",
            "CPATH=/usr/local/cuda/include:/usr/local/cuda/include/cccl:",
            "TORCH_CUDA_ARCH_LIST=8.0;8.6;8.9;9.0;9.0a;10.0;12.0+PTX",
            "FLASHINFER_VERSION=v0.6.12",
            "CUDA_MAJOR=13",
            "CUDA_MINOR=0",
            "LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/lib64:/opt/amazon/efa/lib:/opt/amazon/efa/lib64:/opt/ucx/lib:/opt/ucx/lib64:/opt/nvshmem-v3.4.5-0/lib:/opt/nvshmem-v3.4.5-0/lib64:/usr/local/lib:/usr/local/lib64:",
            "LLM_D_MODELS_DIR=/var/lib/llm-d/models",
            "HF_HOME=/var/lib/llm-d/.hf",
            "HOME=/home/vllm",
            "VLLM_USAGE_SOURCE=production-docker-image",
            "VLLM_WORKER_MULTIPROC_METHOD=fork",
            "OUTLINES_CACHE_DIR=/tmp/outlines",
            "NUMBA_CACHE_DIR=/tmp/numba",
            "TRITON_CACHE_DIR=/tmp/triton",
            "TRITON_LIBCUDA_PATH=/usr/lib64",
            "TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC=15",
            "TORCH_NCCL_DUMP_ON_TIMEOUT=0",
            "VLLM_SKIP_P2P_CHECK=1",
            "VLLM_CACHE_ROOT=/tmp/vllm",
            "UCX_MEM_MMAP_HOOK_MODE=none"
        ],
        "Cmd": null,
        "Image": "",
        "Volumes": null,
        "WorkingDir": "/home/vllm",
        "Entrypoint": [
            "/opt/nvidia/nvidia_entrypoint.sh"
        ],
        "OnBuild": null,
        "Labels": {
            "architecture": "x86_64",
            "build-date": "2025-10-13T07:36:46Z",
            "com.redhat.component": "ubi9-container",
            "com.redhat.license_terms": "https://www.redhat.com/en/about/red-hat-end-user-license-agreements#UBI",
            "cpe": "cpe:/a:redhat:enterprise_linux:9::appstream",
            "description": "The Universal Base Image is designed and engineered to be the base layer for all of your containerized applications, middleware and utilities. This base image is freely redistributable, but Red Hat only supports Red Hat technologies through subscriptions for Red Hat products. This image is maintained by Red Hat and updated regularly.",
            "distribution-scope": "public",
            "io.buildah.version": "1.41.4",
            "io.k8s.description": "The Universal Base Image is designed and engineered to be the base layer for all of your containerized applications, middleware and utilities. This base image is freely redistributable, but Red Hat only supports Red Hat technologies through subscriptions for Red Hat products. This image is maintained by Red Hat and updated regularly.",
            "io.k8s.display-name": "Red Hat Universal Base Image 9",
            "io.openshift.expose-services": "",
            "io.openshift.tags": "base rhel9",
            "maintainer": "NVIDIA CORPORATION \u003csw-cuda-installer@nvidia.com\u003e",
            "name": "ubi9/ubi",
            "org.opencontainers.image.created": "2026-06-24T21:21:44.196Z",
            "org.opencontainers.image.description": "Achieve state of the art inference performance with modern accelerators on Kubernetes",
            "org.opencontainers.image.licenses": "Apache-2.0",
            "org.opencontainers.image.revision": "6afd09cc686309e3fd721cd9d88db8654c2c079b",
            "org.opencontainers.image.source": "https://github.com/llm-d/llm-d",
            "org.opencontainers.image.title": "llm-d",
            "org.opencontainers.image.url": "https://github.com/llm-d/llm-d",
            "org.opencontainers.image.version": "v0.8.0",
            "release": "1760340943",
            "summary": "Provides the latest release of Red Hat Universal Base Image 9.",
            "url": "https://catalog.redhat.com/en/search?searchType=containers",
            "vcs-ref": "60d587d1286655c5e777e63959aaae224123ea95",
            "vcs-type": "git",
            "vendor": "Red Hat, Inc.",
            "version": "9.6"
        }
    },
    "Architecture": "amd64",
    "Os": "linux",
    "Size": 16975827644,
    "GraphDriver": {
        "Data": {
            "LowerDir": "/var/lib/docker/overlay2/6fb60494a436ff3611f2a37e8f9080ca20ced3e5facb834ac0cdd5e86d48e075/diff:/var/lib/docker/overlay2/beb1c30581875430f9e671e31c2a546256c1ad4d405684e55ed2d72628d2f6ab/diff:/var/lib/docker/overlay2/4991acd2b16d43b362e807e2def99251ca49344f6b5d21a9f27248745f77e8f3/diff:/var/lib/docker/overlay2/b79797dbc940ebb039a657449ad2d599f50a14a67a2cbb828010dd5eb6ce9755/diff:/var/lib/docker/overlay2/cb08831c693f50c9cdb9bdc9f1e06ced365d92a3ea44dc4d5f45490a27654624/diff:/var/lib/docker/overlay2/15de908d11ff8b0e219a924e0fe9eaad0dfe0e8364913c23063e232fc45984b9/diff:/var/lib/docker/overlay2/fe7f7bfebe0e37491f25ca941c74aaa663e21c0b91531b89cb490ecd8577358a/diff:/var/lib/docker/overlay2/02ab60af02ef493d9f607ae405c48905691238bc9702416a31dfd248f39a1166/diff:/var/lib/docker/overlay2/4d3ae962595ef2360e746bceabd810615011a53f74d77c9c05f85aab839b04de/diff:/var/lib/docker/overlay2/4c003e8c0f1114d6ea53f133cb74b68d1b89c1a22140abe7b832bf3f9562965d/diff:/var/lib/docker/overlay2/2cc8c37ebdd0b83ee22c320069da31ccf6c9bb66b68d1c6d41ed895c49ea0227/diff:/var/lib/docker/overlay2/9ac1b837cfa6586a3094d4b8c430e3f30329e2a992d7b0009839fe5007dbe2bf/diff:/var/lib/docker/overlay2/c020a00e99318a41d11c7b5f07b242fcb24230fae8485d978b9240db965760ee/diff:/var/lib/docker/overlay2/5bbcd8293611e3c93035b799b013958861c0e8e58358b451711a065d63314608/diff:/var/lib/docker/overlay2/1d53c26ddae5ee0e57d3e3c1c1bbd8ec049afa952451fd7fef874624ddfc6e91/diff:/var/lib/docker/overlay2/a22a3122bd827bca270857f92d531e1a8f45689a8e16f5a9af0dee99d321dc75/diff:/var/lib/docker/overlay2/d3549aa7608271f5f83fa5597bf84f209e78b47cd23fc9931a2ee3c2e860c141/diff:/var/lib/docker/overlay2/2a94e9ec32aecb9498e7756ed1fa8da93c94943accde2ad9107694a25cf3b506/diff:/var/lib/docker/overlay2/dcbc0b3f61bc15054dc211cba951255d22961a716b78233de4ee53797d16f503/diff:/var/lib/docker/overlay2/00e971e0c2ab0ae84ff312d65d5558298fb74b96f1a1da1978e4d0aa4612f15b/diff:/var/lib/docker/overlay2/02a7317d21a069044b0b0dc23232c4638b1c78748d8f388993ae1fbce7b05d05/diff:/var/lib/docker/overlay2/c77018d3d8feb0be12af3d744aa5578a5f4c99753999fd758069e66f282b3c85/diff:/var/lib/docker/overlay2/54c7615869da6d1de6d945f63a8d622b331f45e290e5d82ef28bb01cb7ee5d18/diff:/var/lib/docker/overlay2/1fa03938cdd1e349933a3b8b73d2d038db61b515d19cd91bcf50835578ed52e7/diff:/var/lib/docker/overlay2/8889068768c34b711e1516e1a931533a3bb55c649a801bca1a79f527c1fad993/diff:/var/lib/docker/overlay2/91fab2b9e332ef4eb7723fb03fe8bd4feed4120900b05c3af2564d25e31794ae/diff:/var/lib/docker/overlay2/0c46a08abd0e95b135ae00ee69319d7c7158d8307115d94821fd19f13676dcc9/diff:/var/lib/docker/overlay2/d6b4827f5b73a3c0b1c2ee0ccd7b988abe0883a15f4863940562f938a2437b0b/diff:/var/lib/docker/overlay2/d326d357d13178b97f63f39d30c90352635e5afdd373cb9edc31b0a923929b11/diff:/var/lib/docker/overlay2/f97237f0aba17596df042dad5636072646783883ec43a0c1a709d00ba37c8f9b/diff:/var/lib/docker/overlay2/2703148810c080bde1b3da46b74882fe9807d4925ad5237864713808270d35f1/diff:/var/lib/docker/overlay2/3acb29e338b1f66aff9d14422a4c96b15ceecc3658ff3b16cb4041d93693937e/diff:/var/lib/docker/overlay2/4f6ab0871bc82316a36e75207db695443919eb23098d6d9a4be71ecbf3bf46df/diff:/var/lib/docker/overlay2/8e71ca5e0ac71b5cc2950771be2ea74c171a74004a220d0e0550bdb4bc1e03fb/diff:/var/lib/docker/overlay2/3f7836ee63e129922532127f4dedc5e71b57aa5e173d8290bb36b6dc3d7e7686/diff:/var/lib/docker/overlay2/f28c94a0cddbe87573e8de67c630cdd5ab4d437b7cc3af7abda080fdb3da44de/diff:/var/lib/docker/overlay2/82ac9dce18a6df07f0834a3d92fed605c1f9992046ed66e18dd06a6c13a566ef/diff:/var/lib/docker/overlay2/578e904e723dc8d2ba97e04b2b1765691d5a2657d7f23d89779c4fca96c4839f/diff:/var/lib/docker/overlay2/3f16551db7528c3ddbb81563e50e81d31ab4894db2cf201c26007504080a3451/diff:/var/lib/docker/overlay2/08a9485aa0a6dfa9b55ee87361b424a8c8281f1f967fb8986780810530c9adc3/diff:/var/lib/docker/overlay2/55bcc9ef85afb54b24ec88bc5235194ccb6198306a27e5ae9b5b16488853d185/diff:/var/lib/docker/overlay2/1b2ef696b8c8067abc582ecca4621f1d8dab4a31e7feabcb688140e8a21c6675/diff:/var/lib/docker/overlay2/fc3645aebf6c8c7719d36f260fddcb6176a9d4589306ada066e421d1fc6fef3f/diff:/var/lib/docker/overlay2/3b269a5d2410004f15a5ca346d6f4d231aa13a3a3dbe93873ac51b36ec742608/diff:/var/lib/docker/overlay2/09573253b4fe12cc861953521daa9054e906f69d467ede9424429c56f28fdc1d/diff:/var/lib/docker/overlay2/0b000fc6de614060c8cc8fe44a43b306a54415f4cc345221ebb3b30d527fe9e7/diff:/var/lib/docker/overlay2/bee1308e6ae5a7ca135240276240cb95e83e04e437340a67481b959b79c5e5c2/diff:/var/lib/docker/overlay2/b16f124b7214945009f6c6002da90f33cc1ad729a66d5d182ac9b78f8d03a095/diff:/var/lib/docker/overlay2/744a778ca4d3459da72db3b3ade27ac288efe4ffc7ac0481391f43af3a564bcd/diff",
            "MergedDir": "/var/lib/docker/overlay2/1790919ccdbd7f7f3c535f21b8f5cef0c3d2c792c926880cb248059bf3af67f9/merged",
            "UpperDir": "/var/lib/docker/overlay2/1790919ccdbd7f7f3c535f21b8f5cef0c3d2c792c926880cb248059bf3af67f9/diff",
            "WorkDir": "/var/lib/docker/overlay2/1790919ccdbd7f7f3c535f21b8f5cef0c3d2c792c926880cb248059bf3af67f9/work"
        },
        "Name": "overlay2"
    },
    "RootFS": {
        "Type": "layers",
        "Layers": [
            "sha256:0f4901ccb7578658aace2064d2e383961722007597c3480301c497c399e66a1d",
            "sha256:48c2deda6582ef6ed634d61766eaec39f17ec79a48273829bb2cdda8d2346b93",
            "sha256:7bbf4cf87de6b8a8a0672f847eed0d0f176f06d020607f01aa865869cc416d19",
            "sha256:c127851da2f0ea1a06610c38838a26906932d2670d64bc5efd7223109a4c6663",
            "sha256:148a3befa8f9c1b2df8f9a2a825434b32b7972f9f18c42dd76a748d781f9443d",
            "sha256:737a928854ea4a0c48cc987c41936cd345ebde0fe8f03fd639431933187b3744",
            "sha256:7374b0fcdbb11d10f78f79e64ad9a678e302dc349cd11eb34b62256ac2c9bf41",
            "sha256:6308980268932115a9328c8884c3c98c34d007014b17c5d509a7eb6bea47a52e",
            "sha256:7b747b1807ae0b4c256046fe537b7a1b0d1786e93425dac25a8db801b84ba827",
            "sha256:7732c108f152f210ad27021f043d1679a200ad6dbcc62b242972e7fbfad33f30",
            "sha256:9aece602b01e32b42e290d678aca97ad7460e686fd6149b2864c513a3e4670a3",
            "sha256:42da77b8a921fd95b88f0c189325bcf1ac717d005ba04dd69144aa63f26fdb54",
            "sha256:1f8c6e4c8e28c10b8d2a58783d423e9032a1dc6fb9470f0a31c1f7918611145c",
            "sha256:8d97a5d0d10561e3d6d8fb75652a9c39bf15fbb0a4802a25a5007913f3dceb07",
            "sha256:bbe45cede9d67a97d24645509b5f48d2034d4a80a23eb639209c5a26a1f893c9",
            "sha256:42da77b8a921fd95b88f0c189325bcf1ac717d005ba04dd69144aa63f26fdb54",
            "sha256:1f8c6e4c8e28c10b8d2a58783d423e9032a1dc6fb9470f0a31c1f7918611145c",
            "sha256:34f61dbb058bc2e6b370a4dcdf8e95b6c74333a55fef920699374926d96a12c0",
            "sha256:2d7f6c0f27df51c9ea2462812fd4f5b629ca7746bcbf9f3f125c5494111f84f2",
            "sha256:37edfe84e246bd1d30013a1d0d951083b154716f4bf2da8a7a9a03226d65a91f",
            "sha256:7eaf737c34c502bba0c4861ef9c82ed578674b8865101b847ac42ffe48b08274",
            "sha256:fcc4511daee9497067067fcf3f33d25ae1da1642ceebea846d628c49da43703f",
            "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef",
            "sha256:865fe0989c0e744a5d09f7c21f24877daa766ebf7d1e114ae770d6dba8961abb",
            "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef",
            "sha256:61f9f4aacae38fe8abbfb12619e5c704875e468357dd70d50615ffc3009fcefc",
            "sha256:9b3bca674ad3eb51271a69615c7007d797d623a9dbead494bfa81cac2b4f098b",
            "sha256:159d96d947761562b5b6952cefb5ca509ed76ad85eb152da3f8bc10c25555a9d",
            "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef",
            "sha256:2de1111112a9d8cc9ba276f4f23696573ae835116b4d5bdb1ca3ce7684c356bd",
            "sha256:9441c2608ea767646a12ef45481ba4df7e99bab4db8370d0d35d897e69e01540",
            "sha256:bb6a5d0ed0829081c100bfa687f7b49c5bfd77793c100241b61e309172697c4b",
            "sha256:761e2237b9d92bba812803f211b03b51fba0e347bbd0293dce337e9f1afd4e11",
            "sha256:484e06f0fdc5ac9e1f5625a3a696d623ae94b9a042545801cd4568548c9a916f",
            "sha256:2d3cd1ffde3cd7ec0aad3e6230f57a8bed06213752016453bea20b3fe01ff009",
            "sha256:23e23a03b151b6f1922aa1e20026c62eb3e78c3194c751c0c13ffc904ff51abb",
            "sha256:f2f804d6fb4d40ce7d758adb17e7f9ac15d26e856f7a2d8c3922bda900c85e4a",
            "sha256:9ea4b5382da03b1a18f0c9f90d7d697fcf89e7f6410824d230fe53ffff0d275c",
            "sha256:64aa450e2960e065871c31174995afc35725dec67d5e9df66e6502b02387589b",
            "sha256:5110ec0e2f51e9a46848fd60081323260c521e81ffe60d2020c677881ed3601c",
            "sha256:697093554b394c4628fcb18c3ca02d1c05423dcc317de167e861b8eb870deffb",
            "sha256:88008af0238189ad58517ab6b9203939a44406cad62acc349d56cb653b11a4e1",
            "sha256:37fa57e082c65cebeee1b73876acb8aa8ed8d53ba9f3293747fd17a30591aa62",
            "sha256:133e1555e6c39d113f719ef8005e1899b590fd9a9f5657bf67fe88c2f26530d7",
            "sha256:b376ff39518bd59628725eb9a6502c30eb7ffdb9b2e506d1e47c29fb105340f5",
            "sha256:95953aac587486af3d682b6ffdfd609ca6570d3b757b76aea04d7afa544e3fff",
            "sha256:c5b2062012a4fdcbd0775a081c7d250915fd1a7b425580cbf757256ac70dddcf",
            "sha256:6cecde199f486a009328b0c584b134a6063168025dd04166bdd4bcf4d9fe8559",
            "sha256:c42dc1ca69d4b37d352e5b4b0729ca26059f14cd9a26ac4e03f472d4d7424708",
            "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef"
        ]
    },
    "Metadata": {
        "LastTagTime": "2026-08-22T01:05:28.870671363+08:00"
    }
}

更多版本

ghcr.io/llm-d/llm-d-cuda:v0.8.0

linux/amd64 ghcr.io16.98GB2026-08-22 01:17

0 点赞

0 收藏镜像 Run 生成 Docker Run 命令 K8s 生成 Kubernetes YAML 回到顶部