k8s pod使用gpu前提
- k8s节点有gpu显卡
- k8s节点安装gpu显卡驱动
- k8s节点docker或containerd运行时使用nvidia-container-runtime
- k8s部署gpu device plugin daemonset
1.安装gpu显卡驱动
查看节点显卡类型
nvidia-smi -L
GPU 0: Tesla V100-SXM2-32GB (UUID: GPU-f2b15a66-0630-5f77-1f17-28abb3854f1c)
# 忘记没安装驱动,用不了上面命令,使用
lspci | grep -i nvidia
00:03.0 3D controller: NVIDIA Corporation Device 1eb8 (rev a1)
00:04.0 3D controller: NVIDIA Corporation Device 1eb8 (rev a1)
根据 1eb8到这个网站查
http://pci-ids.ucw.cz/mods/PC/10de?action=help?help=pci
根据型号到这个网站查找驱动安装程序
https://www.nvidia.com/Download/Find.aspx#
# 下载
wget https://us.download.nvidia.com/tesla/515.65.01/NVIDIA-Linux-x86_64-515.65.01.run
chmod +x NVIDIA-Linux-x86_64-515.65.01.run
# 上述安装程序依赖这些包,安装
apt install gcc linux-kernel-headers dkms
sh NVIDIA-Linux-x86_64-515.65.01.run --ui=none --disable-nouveau --no-install-libglvnd --dkms -s
# 使用下面命令验证是否安装成功。
nvidia-smi
Thu Nov 3 19:17:50 2022
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 515.65.01 Driver Version: 515.65.01 CUDA Version: 11.7 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 Tesla V100-SXM2... On | 00000000:00:08.0 Off | 0 |
| N/A 36C P0 37W / 300W | 4MiB / 32768MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| 0 N/A N/A 1499 G /usr/lib/xorg/Xorg 4MiB |
+-----------------------------------------------------------------------------+
2. 安装nvidia-container-runtime
curl -s -L https://nvidia.github.io/nvidia-container-runtime/gpgkey | sudo apt-key add -
curl -s -L https://nvidia.github.io/nvidia-container-runtime/$(. /etc/os-release;echo $ID$VERSION_ID)/nvidia-container-runtime.list | sudo tee /etc/apt/sources.list.d/nvidia-container-runtime.list
apt update
apt install nvidia-container-runtime -y
2.1修改默认运行时
2.2 cri为docker
修改/etc/docker/daemon.json,增加default-runtime,runtimes配置.
{
"default-runtime": "nvidia",
"runtimes": {
"nvidia": {
"path": "/usr/bin/nvidia-container-runtime",
"runtimeArgs": []
}
},
"registry-mirrors": [
"https://docker.mirrors.ustc.edu.cn/",
"https://hub-mirror.c.163.com/"
],
"max-concurrent-downloads": 10,
"log-driver": "json-file",
"log-level": "warn",
"log-opts": {
"max-size": "10m",
"max-file": "3"
},
"insecure-registries":
["127.0.0.1","192.168.12.12:8888"],
"data-root":"/data/docker",
"features":{"buildkit": true}
}
重启
systemctl restart docker
2.2 cri为containerd
修改/etc/containerd/config.toml,如果文件不存在
生成默认配置文件
mkdir /etc/containerd
containerd config default > /etc/containerd/config.toml
vi /etc/containerd/config.toml
...
[plugins."io.containerd.grpc.v1.cri".containerd]
snapshotter = "overlayfs"
default_runtime_name = "runc"
no_pivot = false
...
[plugins."io.containerd.grpc.v1.cri".containerd.runtimes]
[plugins."io.containerd.grpc.v1.cri".containerd.runtimes.runc]
runtime_type = "io.containerd.runtime.v1.linux" # 将此处 runtime_type 的值改成 io.containerd.runtime.v1.linux
...
[plugins."io.containerd.runtime.v1.linux"]
shim = "containerd-shim"
runtime = "nvidia-container-runtime" # 将此处 runtime 的值改成 nvidia-container-runtime
...
重启
systemctl restart containerd
3. 部署nivdia-device-plugin
kubectl apply -f nvidia-device-plugin.yaml
# Copyright (c) 2019, NVIDIA CORPORATION. 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.
apiVersion: apps/v1
kind: DaemonSet
metadata:
name: nvidia-device-plugin-daemonset
namespace: kube-system
spec:
selector:
matchLabels:
name: nvidia-device-plugin-ds
updateStrategy:
type: RollingUpdate
template:
metadata:
# This annotation is deprecated. Kept here for backward compatibility
# See https://kubernetes.io/docs/tasks/administer-cluster/guaranteed-scheduling-critical-addon-pods/
annotations:
scheduler.alpha.kubernetes.io/critical-pod: ""
labels:
name: nvidia-device-plugin-ds
spec:
tolerations:
# This toleration is deprecated. Kept here for backward compatibility
# See https://kubernetes.io/docs/tasks/administer-cluster/guaranteed-scheduling-critical-addon-pods/
- key: CriticalAddonsOnly
operator: Exists
- key: nvidia.com/gpu
operator: Exists
effect: NoSchedule
# Mark this pod as a critical add-on; when enabled, the critical add-on
# scheduler reserves resources for critical add-on pods so that they can
# be rescheduled after a failure.
# See https://kubernetes.io/docs/tasks/administer-cluster/guaranteed-scheduling-critical-addon-pods/
priorityClassName: "system-node-critical"
containers:
- image: nvidia/k8s-device-plugin:v0.7.1
name: nvidia-device-plugin-ctr
args: ["--fail-on-init-error=false"]
securityContext:
allowPrivilegeEscalation: false
capabilities:
drop: ["ALL"]
volumeMounts:
- name: device-plugin
mountPath: /var/lib/kubelet/device-plugins
volumes:
- name: device-plugin
hostPath:
path: /var/lib/kubelet/device-plugins
4. 查看kubelet是否识别gpu
查看pod是否正常启动
kubectl get pod -n kube-system -o wide
describe node查看是否识别gpu
kubectl describe node vm-1-5-ubuntu
测试gpu pod启动
kubectl apply -f gpu-pod.yaml
apiVersion: v1
kind: Pod
metadata:
name: gpu-pod
spec:
containers:
- name: cuda-container
image: nvidia/cuda:9.0-devel
resources:
limits:
nvidia.com/gpu: 1
- name: digits-container
image: nvidia/digits:6.0
resources:
limits:
nvidia.com/gpu: 1
5. gpu共享
以上方式为独占gpu,GPU资源在节点上是以个数暴露给kubernetes集群来进行调度的,也就是说如果有两个后端应用pod需要使用到GPU资源,但集群节点上只有一张GPU物理卡的情况下,会导致两个后端应用容器中仅有一个可以正常运行,另一个pod则会处于pending状态。文章来源:https://www.toymoban.com/news/detail-414380.html
gpu共享的配置阿里云和ucloud等都有对应文档配置,非云集群网上也有开源解决方案。文章来源地址https://www.toymoban.com/news/detail-414380.html
到了这里,关于k8spod使用gpu的文章就介绍完了。如果您还想了解更多内容,请在右上角搜索TOY模板网以前的文章或继续浏览下面的相关文章,希望大家以后多多支持TOY模板网!