MindX DL（昇腾深度学习组件）是反对 Atlas 800 训练服务器、Atlas 800 推理服务器的深度学习组件参考设计，提供昇腾 AI 处理器资源管理和监控、昇腾 AI 处理器优化调度、分布式训练汇合通信配置生成等根底性能，疾速使能合作伙伴进行深度学习平台开发。
操作系统应用的是Ubuntu-1804，CPU是华为自研ARM架构。
一、装置前筹备
- 配置apt网络源
hello@ubuntu:/etc/apt$ sudo cp sources.list~ sources.listhello@ubuntu:/etc/apt$ cat sources.list# # deb cdrom:[Ubuntu-Server 18.04.5 LTS _Bionic Beaver_ - Release arm64 (20200810)]/ bionic main restricted#deb cdrom:[Ubuntu-Server 18.04.5 LTS _Bionic Beaver_ - Release arm64 (20200810)]/ bionic main restricted# See http://help.ubuntu.com/community/UpgradeNotes for how to upgrade to# newer versions of the distribution.deb http://cn.ports.ubuntu.com/ubuntu-ports/ bionic main restricted# deb-src http://cn.ports.ubuntu.com/ubuntu-ports/ bionic main restricted## Major bug fix updates produced after the final release of the## distribution.deb http://cn.ports.ubuntu.com/ubuntu-ports/ bionic-updates main restricted# deb-src http://cn.ports.ubuntu.com/ubuntu-ports/ bionic-updates main restricted## N.B. software from this repository is ENTIRELY UNSUPPORTED by the Ubuntu## team. Also, please note that software in universe WILL NOT receive any## review or updates from the Ubuntu security team.deb http://cn.ports.ubuntu.com/ubuntu-ports/ bionic universe# deb-src http://cn.ports.ubuntu.com/ubuntu-ports/ bionic universedeb http://cn.ports.ubuntu.com/ubuntu-ports/ bionic-updates universe# deb-src http://cn.ports.ubuntu.com/ubuntu-ports/ bionic-updates universe## N.B. software from this repository is ENTIRELY UNSUPPORTED by the Ubuntu ## team, and may not be under a free licence. Please satisfy yourself as to ## your rights to use the software. Also, please note that software in ## multiverse WILL NOT receive any review or updates from the Ubuntu## security team.deb http://cn.ports.ubuntu.com/ubuntu-ports/ bionic multiverse# deb-src http://cn.ports.ubuntu.com/ubuntu-ports/ bionic multiversedeb http://cn.ports.ubuntu.com/ubuntu-ports/ bionic-updates multiverse# deb-src http://cn.ports.ubuntu.com/ubuntu-ports/ bionic-updates multiverse## N.B. software from this repository may not have been tested as## extensively as that contained in the main release, although it includes## newer versions of some applications which may provide useful features.## Also, please note that software in backports WILL NOT receive any review## or updates from the Ubuntu security team.deb http://cn.ports.ubuntu.com/ubuntu-ports/ bionic-backports main restricted universe multiverse# deb-src http://cn.ports.ubuntu.com/ubuntu-ports/ bionic-backports main restricted universe multiverse## Uncomment the following two lines to add software from Canonical's## 'partner' repository.## This software is not part of Ubuntu, but is offered by Canonical and the## respective vendors as a service to Ubuntu users.# deb http://archive.canonical.com/ubuntu bionic partner# deb-src http://archive.canonical.com/ubuntu bionic partnerdeb http://ports.ubuntu.com/ubuntu-ports bionic-security main restricted# deb-src http://ports.ubuntu.com/ubuntu-ports bionic-security main restricteddeb http://ports.ubuntu.com/ubuntu-ports bionic-security universe# deb-src http://ports.ubuntu.com/ubuntu-ports bionic-security universedeb http://ports.ubuntu.com/ubuntu-ports bionic-security multiverse# deb-src http://ports.ubuntu.com/ubuntu-ports bionic-security multiverse2.配置kubernetes网络源
root@ubuntu:~/123/offline-pkg-arm64# cat <<EOF >/etc/apt/sources.list.d/kubernetes.list> deb https://mirrors.aliyun.com/kubernetes/apt/ kubernetes-xenial main> EOF3.创立目录并下载根底包
root@ubuntu:~/123# mkdir offline-pkg-arm64root@ubuntu:~/123# cd offline-pkg-arm64/root@ubuntu:~/123/offline-pkg-arm64# sudo apt updateroot@ubuntu:~/123/offline-pkg-arm64# apt-get download conntrack cri-tools haveged keyutils libhavege1 libltdl7 libnfsidmap2 libtirpc-dev libtirpc1 nfs-common nfs-kernel-server rpcbind socat sshpassroot@ubuntu:~/123/offline-pkg-arm64# wget --no-check-certificate https://download.docker.com/linux/ubuntu/dists/bionic/pool/stable/arm64/docker-ce_18.06.3~ce~3-0~ubuntu_arm64.debroot@ubuntu:~/123/offline-pkg-arm64# apt-get download kubelet=1.17.3-00 kubeadm=1.17.3-00 kubectl=1.17.3-00 kubernetes-cni=0.8.6-004.下载docker镜像并导出保留
root@ubuntu:~/123# mkdir docker_imagesroot@ubuntu:~/123# cd docker_images/root@ubuntu:~/123/docker_images# docker pull calico/node:v3.11.3root@ubuntu:~/123/docker_images# docker save -o calico-node_arm64.tar.gz calico/node:v3.11.3root@ubuntu:~/123/docker_images# docker pull calico/pod2daemon-flexvol:v3.11.3root@ubuntu:~/123/docker_images# docker save -o calico-pod2daemon-flexvol_arm64.tar.gz calico/pod2daemon-flexvol:v3.11.3root@ubuntu:~/123/docker_images# docker pull calico/cni:v3.11.3root@ubuntu:~/123/docker_images# docker save -o calico-cni_arm64.tar.gz calico/cni:v3.11.3root@ubuntu:~/123/docker_images# docker pull calico/kube-controllers:v3.11.3root@ubuntu:~/123/docker_images# docker save -o calico-kube-controllers_arm64.tar.gz calico/kube-controllers:v3.11.3root@ubuntu:~/123/docker_images# docker pull coredns/coredns:1.6.5root@ubuntu:~/123/docker_images# docker save -o coredns_arm64.tar.gz coredns/coredns:1.6.5root@ubuntu:~/123/docker_images# docker pull cruse/etcd-arm64:3.4.3-0root@ubuntu:~/123/docker_images# docker save -o etcd_arm64.tar.gz cruse/etcd-arm64:3.4.3-0root@ubuntu:~/123/docker_images# docker pull cruse/kube-apiserver-arm64:v1.17.3root@ubuntu:~/123/docker_images# docker save -o kube-apiserver_arm64.tar.gz cruse/kube-apiserver-arm64:v1.17.3root@ubuntu:~/123/docker_images# docker pull cruse/kube-controller-manager-arm64:v1.17.3root@ubuntu:~/123/docker_images# docker save -o kube-controller-manager_arm64.tar.gz cruse/kube-controller-manager-arm64:v1.17.3root@ubuntu:~/123/docker_images# docker pull cruse/kube-proxy-arm64:v1.17.3-beta.0root@ubuntu:~/123/docker_images# docker save -o kube-proxy_arm64.tar.gz cruse/kube-proxy-arm64:v1.17.3-beta.0root@ubuntu:~/123/docker_images# docker pull cruse/kube-scheduler-arm64:v1.17.3-beta.0root@ubuntu:~/123/docker_images# docker save -o kube-scheduler_arm64.tar.gz cruse/kube-scheduler-arm64:v1.17.3-beta.0root@ubuntu:~/123/docker_images# docker pull cruse/pause-arm64:3.1root@ubuntu:~/123/docker_images# docker save -o pause_arm64.tar.gz cruse/pause-arm64:3.1root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# docker login -u 15648907522 -p RtZOXgmpYAQd5cj93uFCabNXUWB7wOftGw4pFdcal4XZH4bf06hvFxTOrYtr1nRao ascendhub.huawei.comroot@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# docker pull ascendhub.huawei.com/public-ascendhub/vc-controller-manager_arm64:v1.0.1-r40root@ubuntu:~/123/docker_images# docker pull ascendhub.huawei.com/public-ascendhub/vc-scheduler_arm64:v1.0.1-r40root@ubuntu:~/123/docker_images# docker pull ascendhub.huawei.com/public-ascendhub/vc-webhook-manager_arm64:v1.0.1-r40root@ubuntu:~/123/docker_images# docker pull ascendhub.huawei.com/public-ascendhub/vc-webhook-manager-base_arm64:v1.0.1-r40root@ubuntu:~/123/docker_images# docker pull ascendhub.huawei.com/public-ascendhub/hccl-controller_arm64:v20.2.0root@ubuntu:~/123/docker_images# docker pull ascendhub.huawei.com/public-ascendhub/ascend-k8sdeviceplugin_arm64:v20.2.0root@ubuntu:~/123/docker_images# docker pull ascendhub.huawei.com/public-ascendhub/cadvisor_arm64:v0.34.0-r40root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# docker tag ascendhub.huawei.com/public-ascendhub/vc-controller-manager_arm64:v1.0.1-r40 volcanosh/vc-controller-manager:v1.0.1-r40root@ubuntu:~/123/docker_images# docker tag ascendhub.huawei.com/public-ascendhub/vc-scheduler_arm64:v1.0.1-r40 volcanosh/vc-scheduler:v1.0.1-r40root@ubuntu:~/123/docker_images# docker tag ascendhub.huawei.com/public-ascendhub/vc-webhook-manager_arm64:v1.0.1-r40 volcanosh/vc-webhook-manager:v1.0.1-r40root@ubuntu:~/123/docker_images# docker tag ascendhub.huawei.com/public-ascendhub/vc-webhook-manager-base_arm64:v1.0.1-r40 volcanosh/vc-webhook-manager-base:v1.0.1-r40root@ubuntu:~/123/docker_images# docker tag ascendhub.huawei.com/public-ascendhub/hccl-controller_arm64:v20.2.0 hccl-controller:v20.2.0root@ubuntu:~/123/docker_images# docker tag ascendhub.huawei.com/public-ascendhub/ascend-k8sdeviceplugin_arm64:v20.2.0 ascend-k8sdeviceplugin:v20.2.0root@ubuntu:~/123/docker_images# docker tag ascendhub.huawei.com/public-ascendhub/cadvisor_arm64:v0.34.0-r40 google/cadvisor:v0.34.0-r40root@ubuntu:~/123/docker_images# docker rmi ascendhub.huawei.com/public-ascendhub/vc-controller-manager_arm64:v1.0.1-r40root@ubuntu:~/123/docker_images# docker rmi ascendhub.huawei.com/public-ascendhub/vc-scheduler_arm64:v1.0.1-r40root@ubuntu:~/123/docker_images# docker rmi ascendhub.huawei.com/public-ascendhub/vc-webhook-manager_arm64:v1.0.1-r40root@ubuntu:~/123/docker_images# docker rmi ascendhub.huawei.com/public-ascendhub/vc-webhook-manager-base_arm64:v1.0.1-r40root@ubuntu:~/123/docker_images# docker rmi ascendhub.huawei.com/public-ascendhub/hccl-controller_arm64:v20.2.0root@ubuntu:~/123/docker_images# docker rmi ascendhub.huawei.com/public-ascendhub/ascend-k8sdeviceplugin_arm64:v20.2.0root@ubuntu:~/123/docker_images# docker rmi ascendhub.huawei.com/public-ascendhub/cadvisor_arm64:v0.34.0-r40root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# root@ubuntu:~/123/docker_images# docker save -o Ascend-K8sDevicePlugin-v20.2.0-arm64-Docker.tar.gz ascend-k8sdeviceplugin:v20.2.0root@ubuntu:~/123/docker_images# docker save -o hccl-controller-v20.2.0-arm64.tar.gz hccl-controller:v20.2.0root@ubuntu:~/123/docker_images# docker save -o huawei-cadvisor-v0.34.0-r40-arm64.tar.gz google/cadvisor:v0.34.0-r40root@ubuntu:~/123/docker_images# docker save -o vc-controller-manager-v1.0.1-r40-arm64.tar.gz volcanosh/vc-controller-manager:v1.0.1-r40root@ubuntu:~/123/docker_images# docker save -o vc-scheduler-v1.0.1-r40-arm64.tar.gz volcanosh/vc-scheduler:v1.0.1-r40root@ubuntu:~/123/docker_images# docker save -o vc-webhook-manager-base-v1.0.1-r40-arm64.tar.gz volcanosh/vc-webhook-manager-base:v1.0.1-r40root@ubuntu:~/123/docker_images# docker save -o vc-webhook-manager-v1.0.1-r40-arm64.tar.gz volcanosh/vc-webhook-manager:v1.0.1-r40注* 其中局部镜像是须要在华为hub外面进行获取权限后进行下载
https://support.huaweicloud.c...\_03\_0047.html
5.实现后的目录
root@ubuntu:~/123# tree.├── docker_images│ ├── Ascend-K8sDevicePlugin-v20.2.0-arm64-Docker.tar.gz│ ├── calico-cni_arm64.tar.gz│ ├── calico-kube-controllers_arm64.tar.gz│ ├── calico-node_arm64.tar.gz│ ├── calico-pod2daemon-flexvol_arm64.tar.gz│ ├── coredns_arm64.tar.gz│ ├── etcd_arm64.tar.gz│ ├── hccl-controller-v20.2.0-arm64.tar.gz│ ├── huawei-cadvisor-v0.34.0-r40-arm64.tar.gz│ ├── kube-apiserver_arm64.tar.gz│ ├── kube-controller-manager_arm64.tar.gz│ ├── kube-proxy_arm64.tar.gz│ ├── kube-scheduler_arm64.tar.gz│ ├── pause_arm64.tar.gz│ ├── vc-controller-manager-v1.0.1-r40-arm64.tar.gz│ ├── vc-scheduler-v1.0.1-r40-arm64.tar.gz│ ├── vc-webhook-manager-base-v1.0.1-r40-arm64.tar.gz│ └── vc-webhook-manager-v1.0.1-r40-arm64.tar.gz├── offline-pkg-arm64│ ├── conntrack_1%3a1.4.4+snapshot20161117-6ubuntu2_arm64.deb│ ├── cri-tools_1.13.0-01_arm64.deb│ ├── docker-ce_18.06.3~ce~3-0~ubuntu_arm64.deb│ ├── haveged_1.9.1-6_arm64.deb│ ├── keyutils_1.5.9-9.2ubuntu2_arm64.deb│ ├── kubeadm_1.17.3-00_arm64.deb│ ├── kubectl_1.17.3-00_arm64.deb│ ├── kubelet_1.17.3-00_arm64.deb│ ├── kubernetes-cni_0.8.6-00_arm64.deb│ ├── libhavege1_1.9.1-6_arm64.deb│ ├── libltdl7_2.4.6-2_arm64.deb│ ├── libnfsidmap2_0.25-5.1_arm64.deb│ ├── libtirpc1_0.2.5-1.2ubuntu0.1_arm64.deb│ ├── libtirpc-dev_0.2.5-1.2ubuntu0.1_arm64.deb│ ├── nfs-common_1%3a1.3.4-2.1ubuntu5.5_arm64.deb│ ├── nfs-kernel-server_1%3a1.3.4-2.1ubuntu5.5_arm64.deb│ ├── rpcbind_0.2.3-0.6ubuntu0.18.04.4_arm64.deb│ ├── socat_1.7.3.2-2ubuntu2_arm64.deb│ └── sshpass_1.06-1_arm64.deb├── offline-pkg-arm64.zip└── yamls ├── ascendplugin-310-v20.2.0.yaml ├── ascendplugin-volcano-v20.2.0.yaml ├── cadvisor-v0.34.0-r40.yaml ├── calico.yaml ├── hccl-controller-v20.2.0.yaml ├── npu-exporter-v20.2.0.yaml └── volcano-v1.0.1-r40.yaml3 directories, 46 filesroot@ubuntu:~/123#注* 其中yamls文件在下方链接中下载
https://gitee.com/ascend/mind...
6.配置免密登陆
root@ubuntu:~# ssh-keygenGenerating public/private rsa key pair.Enter file in which to save the key (/root/.ssh/id_rsa): Created directory '/root/.ssh'.Enter passphrase (empty for no passphrase): Enter same passphrase again: Your identification has been saved in /root/.ssh/id_rsa.Your public key has been saved in /root/.ssh/id_rsa.pub.The key fingerprint is:SHA256:07dTbsAycQqT2w7HdCwjIyJig5T20FQ/eHZGxWg7pbY root@ubuntuThe key's randomart image is:+---[RSA 2048]----+| .+... .+. ||o+ . o .+ + ||+o+ ...=BoO + ||...o .o.+/ O || S @ + . || E + = || . o o || o || |+----[SHA256]-----+root@ubuntu:~# root@ubuntu:~# ssh-copy-id -i 127.0.0.17.配置装置ansible
root@ubuntu:~# root@ubuntu:~# apt install ansibleroot@ubuntu:~# vim /etc/ansible/hosts#配置内容如下[all:vars]# default shared directory, you can change it as yoursnfs_shared_dir=/data/atlas_dls# NFS service IPnfs_service_ip=192.168.1.110# Master IPmaster_ip=192.168.1.110# dls install package dirdls_root_dir=/root/123# set proxyproxy=""# Command for logging in to the Asend hubascendhub_login_command="login_command"# Generally, you do not need to change the value or delete it.ascendhub_prefix="ascendhub.huawei.com/public-ascendhub"# versionsdeviceplugin_version="v20.2.0"cadvisor_version="v0.34.0-r40"volcano_version="v1.0.1-r40"hccl_version="v20.2.0"[nfs_server]ubuntu ansible_host=192.168.1.110 ansible_ssh_user="root" ansible_ssh_pass="123123"[localnode]ubuntu ansible_host=192.168.1.110 ansible_ssh_user="root" ansible_ssh_pass="123123"[training_node]ubuntu ansible_host=192.168.1.110 ansible_ssh_user="root" ansible_ssh_pass="123123"[inference_node][A300T_node][arm]ubuntu ansible_host=192.168.1.110 ansible_ssh_user="root" ansible_ssh_pass="123123"[x86][workers:children]training_nodeinference_nodeA300T_noderoot@ubuntu:~/mindxdl/deploy/offline/steps# vim /etc/ansible/ansible.cfglog_path = /var/log/ansible.loghost_key_checking = Falsedeprecation_warnings = False注* 参数阐明，请依据理论写入：
nfs-host-ip：NFS节点服务器IP地址，即服务器IP地址，如果不装置NFS可设置为空字符串，如：""。master-host-ip：治理节点服务器IP地址，即服务器IP地址。install_dir：根底软件包、镜像包和yamls文件夹的上传目录。proxy_address：代理地址，请依据理论状况配置，如果不须要代理，设置为空字符串，如：""。login_command：从Ascend Hub核心获取镜像须要应用的登录命令，仅在线装置须要配置，如："docker login -u xxxxxx@xxxxxx -p xxxxxxxx ascendhub.huawei.com"，留神不要脱漏命令前后的引号，获取形式请参见获取MindX DL镜像中1~2。离线装置可设置为空字符串，如：""。single-node-host-name：请应用单节点主机名，可通过hostname命令查看。IP：服务器IP地址。username：登录服务器的用户名。倡议应用root用户，防止权限有余。passwd：登录服务器的用户明码。二、一键装置
root@ubuntu:~/sshpass# apt install sshpassroot@ubuntu:~/mindxdl/deploy/offline/steps# dos2unix *root@ubuntu:~/mindxdl/deploy/offline/steps# chmod 500 entry.shroot@ubuntu:~/mindxdl/deploy/offline/steps# bash -x entry.sh三、装置后进行验证
1.docker信息查看
root@ubuntu:~# docker infoContainers: 35 Running: 30 Paused: 0 Stopped: 5Images: 18Server Version: 18.06.3-ceStorage Driver: overlay2 Backing Filesystem: extfs Supports d_type: true Native Overlay Diff: trueLogging Driver: json-fileCgroup Driver: systemdPlugins: Volume: local Network: bridge host macvlan null overlay Log: awslogs fluentd gcplogs gelf journald json-file logentries splunk syslogSwarm: inactiveRuntimes: ascend runcDefault Runtime: ascendInit Binary: docker-initcontainerd version: 468a545b9edcd5932818eb9de8e72413e616e86erunc version: a592beb5bc4c4092b1b1bac971afed27687340c5init version: fec3683Security Options: apparmor seccomp Profile: defaultKernel Version: 4.15.0-112-genericOperating System: Ubuntu 18.04.5 LTSOSType: linuxArchitecture: aarch64CPUs: 192Total Memory: 503.6GiBName: ubuntuID: MUTU:QOYU:2P6F:P2QB:4JKZ:QNKE:PPMQ:PQLL:3PDG:QEYU:LMDK:KNMFDocker Root Dir: /var/lib/dockerDebug Mode (client): falseDebug Mode (server): falseRegistry: https://index.docker.io/v1/Labels:Experimental: falseInsecure Registries: docker.mirrors.ustc.edu.cn 127.0.0.0/8Registry Mirrors: https://dockerhub.azk8s.cn/ https://docker.mirrors.ustc.edu.cn/ http://hub-mirror.c.163.com/Live Restore Enabled: falseWARNING: No swap limit support2. kubectl的pod信息查看
root@ubuntu:~# kubectl get pod --all-namespacesNAMESPACE NAME READY STATUS RESTARTS AGEcadvisor cadvisor-nsn4r 1/1 Running 0 5m23sdefault hccl-controller-645bb466f-5fqq6 1/1 Running 0 5m34skube-system ascend-device-plugin-daemonset-vxj8s 1/1 Running 0 5m23skube-system calico-kube-controllers-8464785d6b-bnjdn 1/1 Running 0 5m50skube-system calico-node-blshl 1/1 Running 0 5m51skube-system coredns-6955765f44-5jr59 1/1 Running 0 5m50skube-system coredns-6955765f44-wbzvz 1/1 Running 0 5m50skube-system etcd-ubuntu 1/1 Running 0 5m43skube-system kube-apiserver-ubuntu 1/1 Running 0 5m43skube-system kube-controller-manager-ubuntu 1/1 Running 0 5m43skube-system kube-proxy-b78fm 1/1 Running 0 5m51skube-system kube-scheduler-ubuntu 1/1 Running 0 5m43svolcano-system volcano-admission-74776688c8-g9p9q 1/1 Running 0 5m31svolcano-system volcano-admission-init-sbktn 0/1 Completed 0 5m31svolcano-system volcano-controllers-6786db54f-vn797 1/1 Running 0 5m31svolcano-system volcano-scheduler-844f9b547b-xxjm7 1/1 Running 0 5m31sroot@ubuntu:~# root@ubuntu:~# kubectl describe node ubuntuName: ubuntuRoles: master,workerLabels: accelerator=huawei-Ascend910 beta.kubernetes.io/arch=arm64 beta.kubernetes.io/os=linux host-arch=huawei-arm kubernetes.io/arch=arm64 kubernetes.io/hostname=ubuntu kubernetes.io/os=linux masterselector=dls-master-node node-role.kubernetes.io/master= node-role.kubernetes.io/worker=worker workerselector=dls-worker-nodeAnnotations: huawei.com/Ascend910: Ascend910-1,Ascend910-2,Ascend910-3,Ascend910-0 kubeadm.alpha.kubernetes.io/cri-socket: /var/run/dockershim.sock node.alpha.kubernetes.io/ttl: 0 projectcalico.org/IPv4Address: 192.168.1.110/24 projectcalico.org/IPv4IPIPTunnelAddr: 10.30.243.192 volumes.kubernetes.io/controller-managed-attach-detach: trueCreationTimestamp: Thu, 05 Aug 2021 16:34:33 +0800Taints: <none>Unschedulable: falseLease: HolderIdentity: ubuntu AcquireTime: <unset> RenewTime: Thu, 05 Aug 2021 16:41:29 +0800Conditions: Type Status LastHeartbeatTime LastTransitionTime Reason Message ---- ------ ----------------- ------------------ ------ ------- NetworkUnavailable False Thu, 05 Aug 2021 16:35:06 +0800 Thu, 05 Aug 2021 16:35:06 +0800 CalicoIsUp Calico is running on this node MemoryPressure False Thu, 05 Aug 2021 16:40:30 +0800 Thu, 05 Aug 2021 16:34:27 +0800 KubeletHasSufficientMemory kubelet has sufficient memory available DiskPressure False Thu, 05 Aug 2021 16:40:30 +0800 Thu, 05 Aug 2021 16:34:27 +0800 KubeletHasNoDiskPressure kubelet has no disk pressure PIDPressure False Thu, 05 Aug 2021 16:40:30 +0800 Thu, 05 Aug 2021 16:34:27 +0800 KubeletHasSufficientPID kubelet has sufficient PID available Ready True Thu, 05 Aug 2021 16:40:30 +0800 Thu, 05 Aug 2021 16:35:19 +0800 KubeletReady kubelet is posting ready status. AppArmor enabledAddresses: InternalIP: 192.168.1.110 Hostname: ubuntuCapacity: cpu: 192 ephemeral-storage: 920422204Ki huawei.com/Ascend910: 4 hugepages-2Mi: 0 memory: 528101392Ki pods: 110Allocatable: cpu: 192 ephemeral-storage: 848261101802 huawei.com/Ascend910: 4 hugepages-2Mi: 0 memory: 527998992Ki pods: 110System Info: Machine ID: 3996e745414f461b9e0e990f6d0b597e System UUID: CD56756C-607E-BD02-EB11-5292EAFB068C Boot ID: adb96127-7fdc-4d84-8867-a13005f9b535 Kernel Version: 4.15.0-112-generic OS Image: Ubuntu 18.04.5 LTS Operating System: linux Architecture: arm64 Container Runtime Version: docker://18.6.3 Kubelet Version: v1.17.3 Kube-Proxy Version: v1.17.3PodCIDR: 10.30.0.0/24PodCIDRs: 10.30.0.0/24Non-terminated Pods: (15 in total) Namespace Name CPU Requests CPU Limits Memory Requests Memory Limits AGE --------- ---- ------------ ---------- --------------- ------------- --- cadvisor cadvisor-nsn4r 500m (0%) 1 (0%) 300Mi (0%) 2000Mi (0%) 6m17s default hccl-controller-645bb466f-5fqq6 500m (0%) 500m (0%) 300Mi (0%) 300Mi (0%) 6m28s kube-system ascend-device-plugin-daemonset-vxj8s 500m (0%) 500m (0%) 500Mi (0%) 500Mi (0%) 6m17s kube-system calico-kube-controllers-8464785d6b-bnjdn 0 (0%) 0 (0%) 0 (0%) 0 (0%) 6m44s kube-system calico-node-blshl 250m (0%) 0 (0%) 0 (0%) 0 (0%) 6m45s kube-system coredns-6955765f44-5jr59 100m (0%) 0 (0%) 70Mi (0%) 170Mi (0%) 6m44s kube-system coredns-6955765f44-wbzvz 100m (0%) 0 (0%) 70Mi (0%) 170Mi (0%) 6m44s kube-system etcd-ubuntu 0 (0%) 0 (0%) 0 (0%) 0 (0%) 6m37s kube-system kube-apiserver-ubuntu 250m (0%) 0 (0%) 0 (0%) 0 (0%) 6m37s kube-system kube-controller-manager-ubuntu 200m (0%) 0 (0%) 0 (0%) 0 (0%) 6m37s kube-system kube-proxy-b78fm 0 (0%) 0 (0%) 0 (0%) 0 (0%) 6m45s kube-system kube-scheduler-ubuntu 100m (0%) 0 (0%) 0 (0%) 0 (0%) 6m37s volcano-system volcano-admission-74776688c8-g9p9q 500m (0%) 500m (0%) 300Mi (0%) 300Mi (0%) 6m25s volcano-system volcano-controllers-6786db54f-vn797 500m (0%) 500m (0%) 300Mi (0%) 300Mi (0%) 6m25s volcano-system volcano-scheduler-844f9b547b-xxjm7 500m (0%) 500m (0%) 300Mi (0%) 300Mi (0%) 6m25sAllocated resources: (Total limits may be over 100 percent, i.e., overcommitted.) Resource Requests Limits -------- -------- ------ cpu 4 (2%) 3500m (1%) memory 2140Mi (0%) 4040Mi (0%) ephemeral-storage 0 (0%) 0 (0%) huawei.com/Ascend910 0 0Events: Type Reason Age From Message ---- ------ ---- ---- ------- Normal NodeHasSufficientMemory 7m10s (x8 over 7m11s) kubelet, ubuntu Node ubuntu status is now: NodeHasSufficientMemory Normal NodeHasNoDiskPressure 7m10s (x7 over 7m11s) kubelet, ubuntu Node ubuntu status is now: NodeHasNoDiskPressure Normal NodeHasSufficientPID 7m10s (x6 over 7m11s) kubelet, ubuntu Node ubuntu status is now: NodeHasSufficientPID Normal Starting 6m37s kubelet, ubuntu Starting kubelet. Normal NodeHasSufficientMemory 6m37s kubelet, ubuntu Node ubuntu status is now: NodeHasSufficientMemory Normal NodeHasNoDiskPressure 6m37s kubelet, ubuntu Node ubuntu status is now: NodeHasNoDiskPressure Normal NodeHasSufficientPID 6m37s kubelet, ubuntu Node ubuntu status is now: NodeHasSufficientPID Normal NodeAllocatableEnforced 6m37s kubelet, ubuntu Updated Node Allocatable limit across pods Normal Starting 6m33s kube-proxy, ubuntu Starting kube-proxy. Normal NodeReady 6m17s kubelet, ubuntu Node ubuntu status is now: NodeReadyroot@ubuntu:~#注* 再此信息中能够看到CPU和加速卡的信息
Capacity: cpu: 192 ephemeral-storage: 920422204Ki huawei.com/Ascend910: 4 hugepages-2Mi: 0 memory: 528101392Ki pods: 110Allocatable: cpu: 192 ephemeral-storage: 848261101802 huawei.com/Ascend910: 4 hugepages-2Mi: 0 memory: 527998992Ki pods: 110**详情能够查看华为官网文档：
**
https://support.huaweicloud.c...