-
各位社区小伙伴好,最近在学习华为云相关产品,有一些基础疑问想请教一下,希望能得到大家的指点。如果有相关的实践经验和文档链接,也欢迎分享,先行谢过!
-
各位社区小伙伴好,最近在学习华为云相关产品,有一些基础疑问想请教一下,希望能得到大家的指点。如果有相关的实践经验和文档链接,也欢迎分享,先行谢过!
-
各位社区小伙伴好,最近在学习华为云相关产品,有一些基础疑问想请教一下,希望能得到大家的指点。如果有相关的实践经验和文档链接,也欢迎分享,先行谢过!
-
各位社区小伙伴好,最近在学习华为云相关产品,有一些基础疑问想请教一下,希望能得到大家的指点。如果有相关的实践经验和文档链接,也欢迎分享,先行谢过!
-
团队开始用容器镜像仓库管理私有镜像,现在担心权限问题:怎么让 CI 只能推送、生产集群只能拉取?组织、命名空间、仓库这几层的关系也一直没理清。
-
各位社区小伙伴好,最近在学习华为云相关产品,有一些基础疑问想请教一下,希望能得到大家的指点。如果有相关的实践经验和文档链接,也欢迎分享,先行谢过!
-
各位社区小伙伴好,最近在学习华为云相关产品,有一些基础疑问想请教一下,希望能得到大家的指点。如果有相关的实践经验和文档链接,也欢迎分享,先行谢过!
-
各位社区小伙伴好,最近在学习华为云相关产品,有一些基础疑问想请教一下,希望能得到大家的指点。如果有相关的实践经验和文档链接,也欢迎分享,先行谢过!
-
团队开始用容器镜像仓库管理私有镜像,现在担心权限问题:怎么让 CI 只能推送、生产集群只能拉取?组织、命名空间、仓库这几层的关系也一直没理清。
-
贵阳一机器,实例ID c768c7a7-9633-47d0-adcf-4ed17a252381 名称notebook-c51aERROR 08-25 09:20:47 [core.py:586] File "/vllm-workspace/LMCache-Ascend/lmcache_ascend/integration/vllm/vllm_v1_adapter.py", line 155, in init_lmcache_engineERROR 08-25 09:20:47 [core.py:586] engine = LMCacheEngineBuilder.get_or_create(ERROR 08-25 09:20:47 [core.py:586] File "/vllm-workspace/LMCache/lmcache/v1/cache_engine.py", line 947, in get_or_createERROR 08-25 09:20:47 [core.py:586] memory_allocator = cls._Create_memory_allocator(config, metadata)ERROR 08-25 09:20:47 [core.py:586] File "/vllm-workspace/LMCache-Ascend/lmcache_ascend/v1/cache_engine.py", line 21, in _ascend_create_memory_allocatorERROR 08-25 09:20:47 [core.py:586] return AscendMixedMemoryAllocator(int(max_local_cpu_size * 1024**3))ERROR 08-25 09:20:47 [core.py:586] File "/vllm-workspace/LMCache-Ascend/lmcache_ascend/v1/memory_management.py", line 69, in __init__ERROR 08-25 09:20:47 [core.py:586] lmc_ops.host_register(self.buffer)ERROR 08-25 09:20:47 [core.py:586] RuntimeError: Unable to pin host memory with error code: -1ERROR 08-25 09:20:47 [core.py:586] Exception raised from halRegisterHostPtr at /vllm-workspace/LMCache-Ascend/csrc/managed_mem.cpp:109 (most recent call first):ERROR 08-25 09:20:47 [core.py:586] frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0xb8 (0xfffc2cf2c908 in /usr/local/python3.10.17/lib/python3.10/site-packages/torch/lib/libc10.so)ERROR 08-25 09:20:47 [core.py:586] frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::string const&) + 0x6c (0xfffc2cedb404 in /usr/local/python3.10.17/lib/python3.10/site-packages/torch/lib/libc10.so)ERROR 08-25 09:20:47 [core.py:586] frame #2: <unknown function> + 0x1abf8 (0xfff9c407abf8 in /vllm-workspace/LMCache-Ascend/lmcache_ascend/c_ops.cpython-310-aarch64-linux-gnu.so)运行 Lmcache-ascend 遇到了上述问题,主要是由于可以 pin 的 host memory 有限制,原因是 CPU 的内存锁定方法存在问题,系统的内存锁定限制过低,且在容器环境下没有权限执行 ulimit -l unlimited 来提升内存锁定限制。同时无法调整服务的配置,放开内存锁定---以下是参考资料调整 containerd 服务的配置,放开内存锁定的限制,具体步骤如下: 修改 containerd 服务配置文件:找到 containerd 服务的配置文件,通常路径为 /usr/lib/systemd/system/containerd.service(不同系统可能路径有差异,可通过 systemctl status containerd 查看服务配置文件路径)。添加内存锁定限制配置:在配置文件的 [Service] 部分,添加 LimitMEMLOCK=infinity 配置项,该配置项用于设置内存锁定的限制为无限制。
-
RuntimeError: Unable to pin host memory with error code: -1 · Issue #5 · LMCache/LMCache-Ascend在跑 LMCACHE-ASCEND 的时候,发现会出现如上的错误主要的解决方式就是:部署实例的时候使用 LimitMEMLOCK 或者通过 ulimit -l 解决上限但是由于 notebook 中没有 root 权限,所以无法通过后者解决;由于无法使用 docker run 语句 和 docker-compose 所以无法通过前者解决;想问一下要怎么解决这个内存限制问题
-
以下是 dockerfile 的文件# # Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved. # # 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. # FROM quay.io/ascend/cann:8.2.rc1-910b-openeuler22.03-py3.11 # Set the user ma-user whose UID is 1000 and the user group ma-group whose GID is 100 USER root RUN default_user=$(getent passwd 1000 | awk -F ':' '{print $1}') || echo "uid: 1000 does not exist" && \ default_group=$(getent group 100 | awk -F ':' '{print $1}') || echo "gid: 100 does not exist" && \ if [ ! -z ${default_user} ] && [ ${default_user} != "ma-user" ]; then \ userdel -r ${default_user}; \ fi && \ if [ ! -z ${default_group} ] && [ ${default_group} != "ma-group" ]; then \ groupdel -f ${default_group}; \ fi && \ groupadd -g 100 ma-group && useradd -d /home/ma-user -m -u 1000 -g 100 -s /bin/bash ma-user && \ chmod -R 750 /home/ma-user ARG PIP_INDEX_URL="https://mirrors.aliyun.com/pypi/simple" ARG COMPILE_CUSTOM_KERNELS=1 ENV COMPILE_CUSTOM_KERNELS=${COMPILE_CUSTOM_KERNELS} RUN yum update -y && \ yum install -y python3-pip git vim wget net-tools gcc gcc-c++ make cmake numactl-devel && \ rm -rf /var/cache/yum RUN pip config set global.index-url ${PIP_INDEX_URL} # Set pip source to a faster mirror RUN pip config set global.index-url https://pypi.tuna.tsinghua.edu.cn/simple WORKDIR /workspace COPY . /workspace/LMCache-Ascend/ # Install vLLM ARG VLLM_REPO=https://githubfast.com/vllm-project/vllm.git ARG VLLM_TAG=v0.9.2 RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /workspace/vllm # In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it. RUN VLLM_TARGET_DEVICE="empty" python3 -m pip install -e /workspace/vllm/ --extra-index https://download.pytorch.org/whl/cpu/ --retries 5 --timeout 30 && \ python3 -m pip uninstall -y triton # Install vLLM-Ascend ARG VLLM_ASCEND_REPO=https://githubfast.com/vllm-project/vllm-ascend.git ARG VLLM_ASCEND_TAG=v0.9.2rc1 RUN git clone --depth 1 $VLLM_ASCEND_REPO --branch $VLLM_ASCEND_TAG /workspace/vllm-ascend RUN cd /workspace/vllm-ascend && \ git apply -p1 /workspace/LMCache-Ascend/docker/kv-connector-v1.diff RUN export PIP_EXTRA_INDEX_URL=https://mirrors.huaweicloud.com/ascend/repos/pypi && \ source /usr/local/Ascend/ascend-toolkit/set_env.sh && \ source /usr/local/Ascend/nnal/atb/set_env.sh && \ export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/Ascend/ascend-toolkit/latest/`uname -i`-linux/devlib && \ python3 -m pip install -v -e /workspace/vllm-ascend/ --extra-index https://download.pytorch.org/whl/cpu/ # Install modelscope (for fast download) and ray (for multinode) RUN python3 -m pip install modelscope ray # Install LMCache ARG LMCACHE_REPO=https://githubfast.com/LMCache/LMCache.git ARG LMCACHE_TAG=v0.3.3 RUN git clone --depth 1 $LMCACHE_REPO --branch $LMCACHE_TAG /workspace/LMCache # our build is based on arm64 RUN sed -i "s/^infinistore$/infinistore; platform_machine == 'x86_64'/" /workspace/LMCache/requirements/common.txt # Install LMCache with retries and timeout RUN export NO_CUDA_EXT=1 && python3 -m pip install -v -e /workspace/LMCache --retries 5 --timeout 30 # Install LMCache-Ascend RUN cd /workspace/LMCache-Ascend && \ source /usr/local/Ascend/ascend-toolkit/set_env.sh && \ source /usr/local/Ascend/nnal/atb/set_env.sh && \ export SOC_VERSION=ASCEND910B3 && \ export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/Ascend/ascend-toolkit/latest/`uname -i`-linux/devlib && \ python3 -m pip install -v --no-build-isolation -e . && \ python3 -m pip cache purge # Switch to user ma-user USER ma-user CMD ["/bin/bash"] 注册的镜像选项如下镜像管理界面创建notebook的参数但是最后创建 notebook 失败了
-
docker镜像push时,单层不超10G,且有部分层状态为pushed,但存在一些层为pushing状态,进度条已拉满,但仍push失败,该如何解决?
-
华为云容器镜像docker登陆后push时仍然提示未登录 docker login -u cn-north-4@XXXXXXXXXX-p XXXXXXXXXXXX swr.cn-north-4.myhuaweicloud.com WARNING! Using --password via the CLI is insecure. Use --password-stdin. WARNING! Your password will be stored unencrypted in /home/crf/.docker/config.json. Configure a credential helper to remove this warning. See https://docs.docker.com/engine/reference/commandline/login/#credential-stores Login Succeeded sudo docker push swr.cn-north-4.myhuaweicloud.com/crf/mynginx:1.0 The push refers to repository [swr.cn-north-4.myhuaweicloud.com/xxxx/mynginx] 825fb68b6033: Preparing 7619c0ba3c92: Preparing 1c1f11fd65d6: Preparing 6b133b4de5e6: Preparing 3d07a4a7eb2a: Preparing 756474215d29: Waiting 8d853c8add5d: Waiting denied: You may not login yet
-
想求助是哪里出错了。
上滑加载中
推荐直播
-
华为云码道Agent集成与鸿蒙实战2026/08/11 周二 19:00-21:00
王一男-华为云码道产品规划专家;李炎-华为云码道产品专家;彭江敏-华为云鸿蒙端云一体化开发专家
本次直播带你解读华为云码道7月份产品新特性、新功能。更有专家演示码道Agent Space × 钉钉机器集成实战,从0到1打通消息通道;码道鸿蒙端云一体化实战,快速搭建员工签到系统。
回顾中 -
华为云开发者AI素养直播课·第五期2026/09/04 周五 16:00-18:00
林华鼎-华为云AI开发者运营负责人;蒋春阳-华为云AI开发者案例开发专家
本期直播内容: AI工具体验营 · 第5-8课连讲。Agent-Team 多智能体协作完成毕业设计实践
回顾中 -
华为云开发者AI素养ClassRoom·第六期2026/09/08 周二 19:00-20:00
樊渊-2026华为软件挑战赛冠军
高手来了:看软挑高手解析二维排样问题—从工业难题到算法突破
回顾中
热门标签