国家建设部防化工程师网站官网,ui和平面设计的区别,易企秀+旗下+网站建设,做网站电销Kubernetes Docker 从零部署一个yolov5检测服务#xff0c;服务基于PaddlePaddle/FastDeploy的服务化部署#xff1b;所有软件从零安装。 文章目录1.说明2.环境3.安装过程 3.1安装 Docker 3.2安装 minikube 3.3安装 Kubectl4.部署过程 4.1 Docker相关 4.2 k8s相关 4.3 启动服…Kubernetes Docker 从零部署一个yolov5检测服务服务基于PaddlePaddle/FastDeploy的服务化部署所有软件从零安装。
文章目录1.说明2.环境3.安装过程· 3.1安装 Docker· 3.2安装 minikube· 3.3安装 Kubectl4.部署过程· 4.1 Docker相关· 4.2 k8s相关· 4.3 启动服务· 4.4 客户端测试五.总结六.引用1.说明
基于k8s的minikube部署一个单节点服务重了解在部署流程和细节。服务基于CPU部署。
2.环境 WSL2 Ubuntu 18.04 kubernetes v1.23.8 Docker 23.0.0 PaddlePaddle/FastDeploy 3.安装过程
· 3.1安装 Docker
#1. 新软件列表和允许使用https
sudo apt-get update
sudo apt-get install ca-certificates curl gnupg lsb-release
#2.添加阿里源的GPG
curl -fsSL https://mirrors.aliyun.com/docker-ce/linux/ubuntu/gpg | sudo gpg --dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg#3.设置阿里源的docker仓库echo deb [arch$(dpkg --print-architecture) signed-by/usr/share/keyrings/docker-archive-keyring.gpg] https://mirrors.aliyun.com/docker-ce/linux/ubuntu $(lsb_release -cs) stable | sudo tee /etc/apt/sources.list.d/docker.list /dev/null#4.安装docker:#4.1更新apt-get
sudo apt-get update#4.2安装最新的docker版本
sudo apt-get install docker-ce docker-ce-cli containerd.io#4.3启动docker
sudo service docker start#4.4查看docker服务状态
sudo service docker status· 3.2安装 minikube
curl -LO https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64
sudo install minikube-linux-amd64 /usr/local/bin/minikube这里启动大约要5分钟
#启动k8s,使用docker为引擎.
minikube start --force --driverdocker --kubernetes-versionv1.23.8验证minikube启动情况
minikube kubectl get ns
#显示一下说明成功
rootDESKTOP-4J64IFF:~# minikube kubectl get ns
NAME STATUS AGE
default Active 8d
kube-node-lease Active 8d
kube-public Active 8d
kube-system Active 8d
kubernetes-dashboard Active 8d· 3.3安装 Kubectl
安装
curl -LO https://storage.googleapis.com/kubernetes-release/release/v1.18.0/bin/linux/amd64/kubectl
chmod x ./kubectl
sudo mv ./kubectl /usr/local/bin/kubectl
kubectl version --client测试
kubectl get pod -A4.部署过程
· 4.1 Docker相关
下载CPU镜像仅支持Paddle/ONNX模型在CPU上进行服务化部署支持的推理后端包括OpenVINO、Paddle Inference和ONNX Runtime。
docker pull registry.baidubce.com/paddlepaddle/fastdeploy:1.0.2-cpu-only-21.10下载部署代码及模型等
#下载部署示例代码
git clone https://github.com/PaddlePaddle/FastDeploy.git
cd FastDeploy/examples/vision/detection/yolov5/serving/#下载yolov5模型文件和测试图片
wget https://gitee.com/paddlepaddle/PaddleDetection/raw/release/2.4/demo/000000014439.jpg
wget https://bj.bcebos.com/paddlehub/fastdeploy/yolov5s.onnx# 将模型放入 models/runtime/1目录下, 并重命名为model.onnx
mv yolov5s.onnx models/runtime/1/model.onnx启动docker测试服务可跳过
######################### 服务端 ##############################
cd FastDeploy/examples/vision/detection/yolov5/serving/
# 手动启动docker
docker run -it --nethost --name fd_serving -v pwd/:/yolov5_serving registry.baidubce.com/paddlepaddle/fastdeploy:x.y.z-cpu-only-21.10 bash
#启动服务
fastdeployserver --model-repository/yolov5_serving/models --backend-configpython,shm-default-byte-size10485760#服务启动成功后 会有以下输出:
#I0928 04:51:15.784517 206 grpc_server.cc:4117] Started GRPCInferenceService at 0.0.0.0:8001
#I0928 04:51:15.785177 206 http_server.cc:2815] Started HTTPService at 0.0.0.0:8000
#I0928 04:51:15.826578 206 http_server.cc:167] Started Metrics Service at 0.0.0.0:8002######################### 客户端 ##############################
cd FastDeploy/examples/vision/detection/yolov5/serving/
#安装端依赖
python3 -m pip install tritonclient[all]
# 发送请求
python3 yolov5_grpc_client.py#发送请求成功后会返回json格式的检测结果并打印输出:
#output_name: detction_result
#{boxes: [[268.48028564453125, 81.05305480957031, 298.69476318359375, 169.43902587890625], [104.73116302490234, 45.66197204589844, 127.58382415771484, 93.44938659667969], [378.9093933105469, 39.75013732910156, 395.6086120605469, 84.24342346191406], [158.552978515625, 80.36149597167969, 199.18576049804688, 168.18191528320312], [414.37530517578125, 90.94805908203125, 506.3218994140625, 280.40521240234375], [364.00341796875, 56.608917236328125, 381.97857666015625, 115.96823120117188], [351.7251281738281, 42.635345458984375, 366.9103088378906, 98.04837036132812], [505.8882751464844, 114.36674499511719, 593.1248779296875, 275.99530029296875], [327.7086181640625, 38.36369323730469, 346.84991455078125, 80.89302062988281], [583.493408203125, 114.53289794921875, 612.3546142578125, 175.87353515625], [186.4706573486328, 44.941375732421875, 199.6645050048828, 61.037628173828125], [169.6158905029297, 48.01460266113281, 178.1415557861328, 60.88859558105469], [25.81019401550293, 117.19969177246094, 59.88878631591797, 152.85012817382812], [352.1452941894531, 46.71272277832031, 381.9460754394531, 106.75212097167969], [1.875, 150.734375, 37.96875, 173.78125], [464.65728759765625, 15.901412963867188, 472.512939453125, 34.11640930175781], [64.625, 135.171875, 84.5, 154.40625], [57.8125, 151.234375, 103.0, 174.15625], [165.890625, 88.609375, 527.90625, 339.953125], [101.40625, 152.5625, 118.890625, 169.140625]], scores: [0.8965693116188049, 0.8695310950279236, 0.8684297800064087, 0.8429877758026123, 0.8358422517776489, 0.8151364326477051, 0.8089362382888794, 0.801361083984375, 0.7947245836257935, 0.7606497406959534, 0.6325908303260803, 0.6139386892318726, 0.5906146764755249, 0.505328893661499, 0.40457233786582947, 0.3460320234298706, 0.33283042907714844, 0.3325657248497009, 0.2594234347343445, 0.25389009714126587], label_ids: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 24, 0, 24, 24, 33, 24], masks: [], contain_masks: False}
编写Dockerfile FROM registry.baidubce.com/paddlepaddle/fastdeploy:1.0.2-cpu-only-21.10## 编写人MAINTAINER xxx# 在docker容器构建时拷贝程序ADD ./FastDeploy-develop /opt/FastDeploy-develop# 容器暴露的端口号需要在容器中运行使用端口号一致EXPOSE 8000EXPOSE 8001# 容器启动之后执行的命令 java -jar ROOT.jarCMD [cd /opt/FastDeploy-develop/examples/vision/detection/yolov5/serving]CMD [fastdeployserver,--model-repository/opt/FastDeploy-develop/examples/vision/detection/yolov5/serving/models,--backend-configpython,shm-default-byte-size10485760]打包images
docker build -t k8s_yolo_demo:latest .· 4.2 k8s相关
生成Deployment yaml
kubectl create deployment demo --imagek8s_yolo_demo:latest -o yaml --dry-runclient demo.yaml
#在本地路径下会生成demo.yaml修改demo.yaml
apiVersion: apps/v1
kind: Deployment
metadata:creationTimestamp: nulllabels:app: demoname: demo
spec:replicas: 1selector:matchLabels:app: demostrategy: {}template:metadata:creationTimestamp: nulllabels:app: demospec:containers:- image: k8s_yolo_demoimagePullPolicy: IfNotPresentname: k8s_yolo_demoresources: {} ports:- containerPort: 8001
status: {}编写server yaml
apiVersion: v1
kind: Service
metadata:name: demo
spec:type: NodePort #这里代表是NodePort类型的,另外还有ingress,LoadBalancer这里映射到本机IPports:- port: 8001 #这里的端口和clusterIP(kubectl describe service service-hello中的IP的port)对应即在集群中所有机器上curl 10.98.166.242:80可访问发布的应用服务。targetPort: 8001 #端口一定要和container暴露出来的端口对应nodejs暴露出来的端口是8081所以这里也应是8081nodePort: 31111 # 所有的节点都会开放此端口30000--32767此端口供外部调用。selector:app: demo #这里选择器一定要选择容器的标签之前写name:kube-node是错的。· 4.3 启动服务
# 首先启动
kubectl apply -f demo.yaml
# 可能会报错 获取不到images
# 这里demo.yaml 中添加 imagePullPolicy: IfNotPresent 已添加过了
# 这里使用cache方式获取images# 1.添加环境变量 使得k8s 使用docker
eval $(minikube docker-env)
# 2.把images添加到k8s的cache中去
minikube cache add k8s_yolo_demo
# 3.从新加载启动
kubectl apply -f demo.yaml
# 4.查看启动的pod 是否running
kubectl get pods
# 5.把images从k8s的cache中删掉
minikube cache delete k8s_yolo_demo# 6.启动server
kubectl apply -f demo_server.yaml如图这里看到pod和server都正常启动了
· 4.4 客户端测试
# 查看本机IP
ifconfig#修改客户端代码中的url
url 192.168.49.1:31111python3 yolov5_grpc_client.py#发送请求成功后会返回json格式的检测结果并打印输出:
#output_name: detction_result
#{boxes: [[268.48028564453125, 81.05305480957031, 298.69476318359375, 169.43902587890625], [104.73116302490234, 45.66197204589844, 127.58382415771484, 93.44938659667969], [378.9093933105469, 39.75013732910156, 395.6086120605469, 84.24342346191406], [158.552978515625, 80.36149597167969, 199.18576049804688, 168.18191528320312], [414.37530517578125, 90.94805908203125, 506.3218994140625, 280.40521240234375], [364.00341796875, 56.608917236328125, 381.97857666015625, 115.96823120117188], [351.7251281738281, 42.635345458984375, 366.9103088378906, 98.04837036132812], [505.8882751464844, 114.36674499511719, 593.1248779296875, 275.99530029296875], [327.7086181640625, 38.36369323730469, 346.84991455078125, 80.89302062988281], [583.493408203125, 114.53289794921875, 612.3546142578125, 175.87353515625], [186.4706573486328, 44.941375732421875, 199.6645050048828, 61.037628173828125], [169.6158905029297, 48.01460266113281, 178.1415557861328, 60.88859558105469], [25.81019401550293, 117.19969177246094, 59.88878631591797, 152.85012817382812], [352.1452941894531, 46.71272277832031, 381.9460754394531, 106.75212097167969], [1.875, 150.734375, 37.96875, 173.78125], [464.65728759765625, 15.901412963867188, 472.512939453125, 34.11640930175781], [64.625, 135.171875, 84.5, 154.40625], [57.8125, 151.234375, 103.0, 174.15625], [165.890625, 88.609375, 527.90625, 339.953125], [101.40625, 152.5625, 118.890625, 169.140625]], scores: [0.8965693116188049, 0.8695310950279236, 0.8684297800064087, 0.8429877758026123, 0.8358422517776489, 0.8151364326477051, 0.8089362382888794, 0.801361083984375, 0.7947245836257935, 0.7606497406959534, 0.6325908303260803, 0.6139386892318726, 0.5906146764755249, 0.505328893661499, 0.40457233786582947, 0.3460320234298706, 0.33283042907714844, 0.3325657248497009, 0.2594234347343445, 0.25389009714126587], label_ids: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 24, 0, 24, 24, 33, 24], masks: [], contain_masks: False}五.总结
最有问题的还是pod IP映射的问题没搞清k8s的端口映射逻辑。本机由于虚拟机启动IP会变动为 192.168.49.1实际服务是之前启动的已经映射到了192.168.49.2上这里需要从新启动下server就会从新映射到当前IP。这里k8s获取docker image的方式也有问题这是测试无所谓了正式部署不会存在这种问题。GPU方式还没尝试有空了再试试主要是搭建环境比较麻烦。
六.引用 https://github.com/PaddlePaddle/FastDeploy/blob/develop/examples/vision/detection/yolov5/serving/README_CN.md https://www.jianshu.com/p/c8b42d5cda2d
