Docker拉取命令

docker pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64
docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64  docker.io/hicasper/llama-cpp-cuda:server-b10068

Containerd拉取命令

ctr images pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64
ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64  docker.io/hicasper/llama-cpp-cuda:server-b10068

Shell快速替换命令

sed -i 's#hicasper/llama-cpp-cuda:server-b10068#swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64#' deployment.yaml

Ansible快速分发-Docker

#ansible k8s -m shell -a 'docker pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64 && docker tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64  docker.io/hicasper/llama-cpp-cuda:server-b10068'

Ansible快速分发-Containerd

#ansible k8s -m shell -a 'ctr images pull swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64 && ctr images tag  swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64  docker.io/hicasper/llama-cpp-cuda:server-b10068'

镜像构建历史

# 2026-07-20 03:16:44  0.00B 配置容器启动时运行的命令
ENTRYPOINT ["/app/llama-server"]
                        
# 2026-07-20 03:16:44  0.00B 指定检查容器健康状态的命令
HEALTHCHECK {Test:[CMD curl -f http://localhost:8080/health] Interval:0s Timeout:0s StartPeriod:0s StartInterval:0s Retries:0}
                        
# 2026-07-20 03:16:44  0.00B 设置工作目录为/app
WORKDIR /app
                        
# 2026-07-20 03:16:44  4.64MB 复制新文件或目录到容器中
COPY /app/full/llama /app/full/llama-server /app # buildkit
                        
# 2026-07-20 03:16:44  0.00B 设置环境变量 LLAMA_ARG_HOST
ENV LLAMA_ARG_HOST=0.0.0.0
                        
# 2026-07-20 03:16:44  168.93MB 复制新文件或目录到容器中
COPY /app/lib/ /app # buildkit
                        
# 2026-07-20 03:07:06  404.57MB 执行命令并创建新的镜像层
RUN |5 BUILD_DATE=2026-07-19T19:05:00Z APP_VERSION=b10068 APP_REVISION= IMAGE_URL=https://github.com/ggml-org/llama.cpp IMAGE_SOURCE=https://github.com/ggml-org/llama.cpp /bin/sh -c apt-get update     && apt-get install -y libgomp1 curl ffmpeg     && apt autoremove -y     && apt clean -y     && rm -rf /tmp/* /var/tmp/*     && find /var/cache/apt/archives /var/lib/apt/lists -not -name lock -type f -delete     && find /var/cache -type f -delete # buildkit
                        
# 2026-07-20 03:07:06  0.00B 添加元数据标签
LABEL org.opencontainers.image.created=2026-07-19T19:05:00Z org.opencontainers.image.version=b10068 org.opencontainers.image.title=llama.cpp org.opencontainers.image.description=LLM inference in C/C++ org.opencontainers.image.url=https://github.com/ggml-org/llama.cpp org.opencontainers.image.source=https://github.com/ggml-org/llama.cpp
                        
# 2026-07-20 03:07:06  0.00B 定义构建参数
ARG IMAGE_SOURCE=https://github.com/ggml-org/llama.cpp
                        
# 2026-07-20 03:07:06  0.00B 定义构建参数
ARG IMAGE_URL=https://github.com/ggml-org/llama.cpp
                        
# 2026-07-20 03:07:06  0.00B 定义构建参数
ARG APP_REVISION=
                        
# 2026-07-20 03:07:06  0.00B 定义构建参数
ARG APP_VERSION=b10068
                        
# 2026-07-20 03:07:06  0.00B 定义构建参数
ARG BUILD_DATE=2026-07-19T19:05:00Z
                        
# 2023-11-10 13:05:11  0.00B 配置容器启动时运行的命令
ENTRYPOINT ["/opt/nvidia/nvidia_entrypoint.sh"]
                        
# 2023-11-10 13:05:11  0.00B 设置环境变量 NVIDIA_PRODUCT_NAME
ENV NVIDIA_PRODUCT_NAME=CUDA
                        
# 2023-11-10 13:05:11  2.53KB 复制新文件或目录到容器中
COPY nvidia_entrypoint.sh /opt/nvidia/ # buildkit
                        
# 2023-11-10 13:05:11  3.06KB 复制新文件或目录到容器中
COPY entrypoint.d/ /opt/nvidia/entrypoint.d/ # buildkit
                        
# 2023-11-10 13:05:11  259.78KB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=arm64 /bin/sh -c apt-mark hold ${NV_LIBCUBLAS_PACKAGE_NAME} ${NV_LIBNCCL_PACKAGE_NAME} # buildkit
                        
# 2023-11-10 13:05:10  1.91GB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=arm64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     cuda-libraries-12-2=${NV_CUDA_LIB_VERSION}     ${NV_LIBNPP_PACKAGE}     cuda-nvtx-12-2=${NV_NVTX_VERSION}     libcusparse-12-2=${NV_LIBCUSPARSE_VERSION}     ${NV_LIBCUBLAS_PACKAGE}     ${NV_LIBNCCL_PACKAGE}     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2023-11-10 13:05:10  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2023-11-10 13:05:10  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_LIBNCCL_PACKAGE
ENV NV_LIBNCCL_PACKAGE=libnccl2=2.19.3-1+cuda12.2
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NCCL_VERSION
ENV NCCL_VERSION=2.19.3-1
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_LIBNCCL_PACKAGE_VERSION
ENV NV_LIBNCCL_PACKAGE_VERSION=2.19.3-1
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_LIBNCCL_PACKAGE_NAME
ENV NV_LIBNCCL_PACKAGE_NAME=libnccl2
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_LIBCUBLAS_PACKAGE
ENV NV_LIBCUBLAS_PACKAGE=libcublas-12-2=12.2.5.6-1
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_LIBCUBLAS_VERSION
ENV NV_LIBCUBLAS_VERSION=12.2.5.6-1
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_LIBCUBLAS_PACKAGE_NAME
ENV NV_LIBCUBLAS_PACKAGE_NAME=libcublas-12-2
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_LIBCUSPARSE_VERSION
ENV NV_LIBCUSPARSE_VERSION=12.1.2.141-1
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_LIBNPP_PACKAGE
ENV NV_LIBNPP_PACKAGE=libnpp-12-2=12.2.1.4-1
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_LIBNPP_VERSION
ENV NV_LIBNPP_VERSION=12.2.1.4-1
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_NVTX_VERSION
ENV NV_NVTX_VERSION=12.2.140-1
                        
# 2023-11-10 13:05:10  0.00B 设置环境变量 NV_CUDA_LIB_VERSION
ENV NV_CUDA_LIB_VERSION=12.2.2-1
                        
# 2023-10-19 02:58:00  0.00B 设置环境变量 NVIDIA_DRIVER_CAPABILITIES
ENV NVIDIA_DRIVER_CAPABILITIES=compute,utility
                        
# 2023-10-19 02:58:00  0.00B 设置环境变量 NVIDIA_VISIBLE_DEVICES
ENV NVIDIA_VISIBLE_DEVICES=all
                        
# 2023-10-19 02:58:00  17.29KB 复制新文件或目录到容器中
COPY NGC-DL-CONTAINER-LICENSE / # buildkit
                        
# 2023-10-19 02:58:00  0.00B 设置环境变量 LD_LIBRARY_PATH
ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64
                        
# 2023-10-19 02:58:00  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
                        
# 2023-10-19 02:58:00  46.00B 执行命令并创建新的镜像层
RUN |1 TARGETARCH=arm64 /bin/sh -c echo "/usr/local/nvidia/lib" >> /etc/ld.so.conf.d/nvidia.conf     && echo "/usr/local/nvidia/lib64" >> /etc/ld.so.conf.d/nvidia.conf # buildkit
                        
# 2023-10-19 02:58:00  1.43MB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=arm64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     cuda-cudart-12-2=${NV_CUDA_CUDART_VERSION}     ${NV_CUDA_COMPAT_PACKAGE}     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2023-10-19 02:58:00  0.00B 设置环境变量 CUDA_VERSION
ENV CUDA_VERSION=12.2.2
                        
# 2023-10-19 02:58:00  10.21MB 执行命令并创建新的镜像层
RUN |1 TARGETARCH=arm64 /bin/sh -c apt-get update && apt-get install -y --no-install-recommends     gnupg2 curl ca-certificates &&     curl -fsSLO https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/${NVARCH}/cuda-keyring_1.0-1_all.deb &&     dpkg -i cuda-keyring_1.0-1_all.deb &&     apt-get purge --autoremove -y curl     && rm -rf /var/lib/apt/lists/* # buildkit
                        
# 2023-10-19 02:58:00  0.00B 添加元数据标签
LABEL maintainer=NVIDIA CORPORATION <cudatools@nvidia.com>
                        
# 2023-10-19 02:58:00  0.00B 定义构建参数
ARG TARGETARCH
                        
# 2023-10-19 02:58:00  0.00B 设置环境变量 NV_CUDA_CUDART_VERSION
ENV NV_CUDA_CUDART_VERSION=12.2.140-1
                        
# 2023-10-19 02:58:00  0.00B 设置环境变量 NVIDIA_REQUIRE_CUDA
ENV NVIDIA_REQUIRE_CUDA=cuda>=12.2
                        
# 2023-10-19 02:58:00  0.00B 设置环境变量 NVARCH
ENV NVARCH=sbsa
                        
# 2023-10-05 15:32:22  0.00B 
/bin/sh -c #(nop)  CMD ["/bin/bash"]
                        
# 2023-10-05 15:32:22  69.19MB 
/bin/sh -c #(nop) ADD file:f8594e26831508c318e42c8dfd9942041031087b8de1bf3fec11fd75b8b30fd4 in / 
                        
# 2023-10-05 15:32:21  0.00B 
/bin/sh -c #(nop)  LABEL org.opencontainers.image.version=22.04
                        
# 2023-10-05 15:32:21  0.00B 
/bin/sh -c #(nop)  LABEL org.opencontainers.image.ref.name=ubuntu
                        
# 2023-10-05 15:32:21  0.00B 
/bin/sh -c #(nop)  ARG LAUNCHPAD_BUILD_ARCH
                        
# 2023-10-05 15:32:20  0.00B 
/bin/sh -c #(nop)  ARG RELEASE
                        
                    

镜像信息

{
    "Id": "sha256:b81479e7d88621774d19c1a2def9276e6d15765b9285444274d6cf8cbfd98b5d",
    "RepoTags": [
        "hicasper/llama-cpp-cuda:server-b10068",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda:server-b10068-linuxarm64"
    ],
    "RepoDigests": [
        "hicasper/llama-cpp-cuda@sha256:2afa8e6a6d598832e3a30187f87e84e847420be5c09a4a470f604c2d8be3ed22",
        "swr.cn-north-4.myhuaweicloud.com/ddn-k8s/docker.io/hicasper/llama-cpp-cuda@sha256:0b573b023b014535fabf3f25667768a1152ce78b8c031c5bf4f1b44954d3fbcb"
    ],
    "Parent": "",
    "Comment": "buildkit.dockerfile.v0",
    "Created": "2026-07-19T19:16:44.190528571Z",
    "Container": "",
    "ContainerConfig": null,
    "DockerVersion": "",
    "Author": "",
    "Config": {
        "Hostname": "",
        "Domainname": "",
        "User": "",
        "AttachStdin": false,
        "AttachStdout": false,
        "AttachStderr": false,
        "Tty": false,
        "OpenStdin": false,
        "StdinOnce": false,
        "Env": [
            "PATH=/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin",
            "NVARCH=sbsa",
            "NVIDIA_REQUIRE_CUDA=cuda\u003e=12.2",
            "NV_CUDA_CUDART_VERSION=12.2.140-1",
            "CUDA_VERSION=12.2.2",
            "LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64",
            "NVIDIA_VISIBLE_DEVICES=all",
            "NVIDIA_DRIVER_CAPABILITIES=compute,utility",
            "NV_CUDA_LIB_VERSION=12.2.2-1",
            "NV_NVTX_VERSION=12.2.140-1",
            "NV_LIBNPP_VERSION=12.2.1.4-1",
            "NV_LIBNPP_PACKAGE=libnpp-12-2=12.2.1.4-1",
            "NV_LIBCUSPARSE_VERSION=12.1.2.141-1",
            "NV_LIBCUBLAS_PACKAGE_NAME=libcublas-12-2",
            "NV_LIBCUBLAS_VERSION=12.2.5.6-1",
            "NV_LIBCUBLAS_PACKAGE=libcublas-12-2=12.2.5.6-1",
            "NV_LIBNCCL_PACKAGE_NAME=libnccl2",
            "NV_LIBNCCL_PACKAGE_VERSION=2.19.3-1",
            "NCCL_VERSION=2.19.3-1",
            "NV_LIBNCCL_PACKAGE=libnccl2=2.19.3-1+cuda12.2",
            "NVIDIA_PRODUCT_NAME=CUDA",
            "LLAMA_ARG_HOST=0.0.0.0"
        ],
        "Cmd": null,
        "Healthcheck": {
            "Test": [
                "CMD",
                "curl",
                "-f",
                "http://localhost:8080/health"
            ]
        },
        "Image": "",
        "Volumes": null,
        "WorkingDir": "/app",
        "Entrypoint": [
            "/app/llama-server"
        ],
        "OnBuild": null,
        "Labels": {
            "maintainer": "NVIDIA CORPORATION \u003ccudatools@nvidia.com\u003e",
            "org.opencontainers.image.created": "2026-07-19T19:05:00Z",
            "org.opencontainers.image.description": "LLM inference in C/C++",
            "org.opencontainers.image.ref.name": "ubuntu",
            "org.opencontainers.image.source": "https://github.com/ggml-org/llama.cpp",
            "org.opencontainers.image.title": "llama.cpp",
            "org.opencontainers.image.url": "https://github.com/ggml-org/llama.cpp",
            "org.opencontainers.image.version": "b10068"
        }
    },
    "Architecture": "arm64",
    "Os": "linux",
    "Size": 2567896122,
    "GraphDriver": {
        "Data": {
            "LowerDir": "/var/lib/docker/overlay2/d6221191ba0b5a6edde3557292d5b7699036ad8773e859c71d1dbf139785a894/diff:/var/lib/docker/overlay2/d8d69755f16c892a88ec91a519840adf8ae1d6ea612a532e3e0db8a1d0f93f1d/diff:/var/lib/docker/overlay2/2b9b8c90810e04512cf778ad4c4ca75d2d35d7d2386c1bf20a238bcb7aa10806/diff:/var/lib/docker/overlay2/8de5809ce5ba4f9835f86643df459b3a166138b686026f28d2dd351e34df7c7b/diff:/var/lib/docker/overlay2/1558747434bad16be4894b0af1aec8c3a8eb5cc4f55d62cf24098dd585fc3fdd/diff:/var/lib/docker/overlay2/eded012f5eda1d80ccc8a7430b15fbd4ba19622957185d903b6d7750816f99e7/diff:/var/lib/docker/overlay2/10f682e174a7f0bcad170e8323e4ec9f2c272d340aefcf3a6b5e79c8a942fadb/diff:/var/lib/docker/overlay2/ecdbb6e12659eee0da8009fcb2a340ab46878b9627bb0657480c8a3d597e0fe4/diff:/var/lib/docker/overlay2/f248d093c30cbc4f7d28925d47c6bbe303684c3dc8a796c1f3cff10c3f5f4d4c/diff:/var/lib/docker/overlay2/db96a3c183fe3b31d844ee07c043a72280ac86b2cebc8b4fc78d6b7cc335a20c/diff:/var/lib/docker/overlay2/b1b8c36cb516303e689fef9b869465259d1e08ddfef2f7e1aa4aa30fe6de2ed6/diff:/var/lib/docker/overlay2/fbfdae97bb508881be2478c91f3ae39180f4580d97b8ac010e4f431dd929813d/diff",
            "MergedDir": "/var/lib/docker/overlay2/927a003a146fa2f012c76c80565f71a028747ed38f282041ed1204a8ecd3b12e/merged",
            "UpperDir": "/var/lib/docker/overlay2/927a003a146fa2f012c76c80565f71a028747ed38f282041ed1204a8ecd3b12e/diff",
            "WorkDir": "/var/lib/docker/overlay2/927a003a146fa2f012c76c80565f71a028747ed38f282041ed1204a8ecd3b12e/work"
        },
        "Name": "overlay2"
    },
    "RootFS": {
        "Type": "layers",
        "Layers": [
            "sha256:d2d3127fc3d362a0516c53fcb8fb983ebb9d1ea2dfbd63b69967a0084f0d45a3",
            "sha256:06bc48fa89a5b6e19e08c842d5a1a65a8ab13e772daacb29bd890a021efa3dc4",
            "sha256:a519aa56714324b172f44446a7eea26c99557a5df134b27ecd4d16493c093871",
            "sha256:dbe62fe3162ed75c2f5d96e33cbf95e1b90a1886ede00da2b804895d8186d877",
            "sha256:6a7da94484fc0e37b5c73da57769d7b96f80de9ffb44bda21f8092439ce5ca35",
            "sha256:4533b5d1a66beef41e21c12d92576508a279c2938a5e70e7dece7f3928511c22",
            "sha256:e962644113c78edfcff5e0169b28f72b7378d7db0be8aa02c245be26836e8b2d",
            "sha256:6c5ea3a312ce26e178d89e3f066219c8ab6ff87281708c7d4d3374c0b7cd84ad",
            "sha256:837e218ad1039000d780b1eb4f4d90414d95c1ebb9e90e8c32557f9a8983b527",
            "sha256:23bb0f3cc6c6cf40b5a75ccc73cfe95d72839ab82ab7fb1776694c57b5977f39",
            "sha256:61ba051f90f9ee733b33357fdcf8724672f9e92a4b33a51ee29609b01cc12cf9",
            "sha256:255723b4132fc7a11e1c835c93c477912c942bf912ca1ec487d19ac534aac85f",
            "sha256:5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef"
        ]
    },
    "Metadata": {
        "LastTagTime": "2026-07-20T03:24:51.4148371+08:00"
    }
}