1 前言

一个问题，一篇文章，一出故事。
由于笔者今天接到部署Python的应用的工作任务，于是整理此文。

2 最佳实践

2.1 编译安装特定的版本

2.1.1 安装Python

如何编译安装Python3？

2.1.2 检查是否满足版本需求

python -V

可见如下显示，

Python 3.9.18

2.2 部署应用

2.2.1 创建运行用户

useradd -d /data/python python
su - python

2.2.2 创建虚拟环境

python -m venv venv

激活虚拟环境

source venv/bin/activate

进入程序目录

unzip cv_flask_web.zip
cd cv_flask_web

2.2.3 安装依赖关系

pip install -r requirements.txt

2.2.4 尝试启动程序

python yolo_webapp.py

如果遇到如下错误提示，

Traceback (most recent call last):
  File "/data/python/cv_flask_web/yolo_webapp.py", line 1, in 
    from flask import Flask,render_template,Response,request
ModuleNotFoundError: No module named 'flask'

你可能需要使用如下命令解决，

pip install flask

如果遇到如下错误提示，

Traceback (most recent call last):
File "/data/python/cv_flask_web/yolo_webapp.py", line 2, in 
import cv2
ModuleNotFoundError: No module named 'cv2'

你可能需要使用如下命令解决，

pip install opencv-python

如果遇到如下错误提示，

Traceback (most recent call last):
  File "/data/python/cv_flask_web/yolo_webapp.py", line 2, in 
    import cv2
  File "/data/python/venv/lib64/python3.9/site-packages/cv2/__init__.py", line 181, in 
    bootstrap()
  File "/data/python/venv/lib64/python3.9/site-packages/cv2/__init__.py", line 153, in bootstrap
    native_module = importlib.import_module("cv2")
  File "/usr/lib64/python3.9/importlib/__init__.py", line 127, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
ImportError: libGL.so.1: cannot open shared object file: No such file or directory

你可能需要使用如下命令解决，

dnf install -y mesa-libGL

如果遇到如下错误提示，

Traceback (most recent call last):
  File "/data/python/cv_flask_web/yolo_webapp.py", line 3, in 
    import torch
ModuleNotFoundError: No module named 'torch'

你可能需要使用如下命令解决，

pip install torch

如果遇到如下错误提示，

Traceback (most recent call last):
  File "/data/python/cv_flask_web/yolo_webapp.py", line 25, in 
    model = torch.hub.load('ultralytics/yolov5', 'custom', path = 'clean_rm_r5.pt')
  File "/data/python/venv/lib64/python3.9/site-packages/torch/hub.py", line 638, in load
    repo_or_dir = _get_cache_or_reload(
  File "/data/python/venv/lib64/python3.9/site-packages/torch/hub.py", line 242, in _get_cache_or_reload
    _check_repo_is_trusted(
  File "/data/python/venv/lib64/python3.9/site-packages/torch/hub.py", line 312, in _check_repo_is_trusted
    Path(filepath).touch()
  File "/usr/lib64/python3.9/pathlib.py", line 1084, in __new__
    raise NotImplementedError("cannot instantiate %r on your system"
NotImplementedError: cannot instantiate 'WindowsPath' on your system

你可能需要使用如下命令解决，

cp /data/python/cv_flask_web/yolo_webapp.py /data/python/cv_flask_web/yolo_webapp.py.save
vim /data/python/cv_flask_web/yolo_webapp.py

修改如下代码，

#pathlib.PosixPath = pathlib.WindowsPath
pathlib.WindowsPath = pathlib.PosixPath

如果遇到如下错误提示，

Using cache found in /data/python/.cache/torch/hub/ultralytics_yolov5_master
Traceback (most recent call last):
  File "/data/python/cv_flask_web/yolo_webapp.py", line 26, in 
    model = torch.hub.load('ultralytics/yolov5', 'custom', path = 'clean_rm_r5.pt')
  File "/data/python/venv/lib64/python3.9/site-packages/torch/hub.py", line 647, in load
    model = _load_local(repo_or_dir, model, *args, **kwargs)
  File "/data/python/venv/lib64/python3.9/site-packages/torch/hub.py", line 676, in _load_local
    model = entry(*args, **kwargs)
  File "/data/python/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py", line 135, in custom
    return _create(path, autoshape=autoshape, verbose=_verbose, device=device)
  File "/data/python/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py", line 54, in _create
    from models.common import AutoShape, DetectMultiBackend
  File "/data/python/.cache/torch/hub/ultralytics_yolov5_master/models/common.py", line 18, in 
    import pandas as pd
ModuleNotFoundError: No module named 'pandas'

你可能需要使用如下命令解决，

pip install pandas

如果遇到如下错误提示，

Using cache found in /data/python/.cache/torch/hub/ultralytics_yolov5_master
Traceback (most recent call last):
  File "/data/python/cv_flask_web/yolo_webapp.py", line 26, in 
    model = torch.hub.load('ultralytics/yolov5', 'custom', path = 'clean_rm_r5.pt')
  File "/data/python/venv/lib64/python3.9/site-packages/torch/hub.py", line 647, in load
    model = _load_local(repo_or_dir, model, *args, **kwargs)
  File "/data/python/venv/lib64/python3.9/site-packages/torch/hub.py", line 676, in _load_local
    model = entry(*args, **kwargs)
  File "/data/python/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py", line 135, in custom
    return _create(path, autoshape=autoshape, verbose=_verbose, device=device)
  File "/data/python/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py", line 54, in _create
    from models.common import AutoShape, DetectMultiBackend
  File "/data/python/.cache/torch/hub/ultralytics_yolov5_master/models/common.py", line 19, in 
    import requests
ModuleNotFoundError: No module named 'requests'

你可能需要使用如下命令解决，

pip install requests

如果遇到如下错误提示，

Using cache found in /data/python/.cache/torch/hub/ultralytics_yolov5_master
Traceback (most recent call last):
  File "/data/python/cv_flask_web/yolo_webapp.py", line 26, in 
    model = torch.hub.load('ultralytics/yolov5', 'custom', path = 'clean_rm_r5.pt')
  File "/data/python/venv/lib64/python3.9/site-packages/torch/hub.py", line 647, in load
    model = _load_local(repo_or_dir, model, *args, **kwargs)
  File "/data/python/venv/lib64/python3.9/site-packages/torch/hub.py", line 676, in _load_local
    model = entry(*args, **kwargs)
  File "/data/python/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py", line 135, in custom
    return _create(path, autoshape=autoshape, verbose=_verbose, device=device)
  File "/data/python/.cache/torch/hub/ultralytics_yolov5_master/hubconf.py", line 54, in _create
    from models.common import AutoShape, DetectMultiBackend
  File "/data/python/.cache/torch/hub/ultralytics_yolov5_master/models/common.py", line 22, in 
    from PIL import Image
ModuleNotFoundError: No module named 'PIL'

你可能需要使用如下命令解决，

pip install Pillow

2.2.5 测试使用gunicorn启动

pip install gunicorn

以上命令安装软件包后，你使用如下命令测试服务启动，

cd cv_flask_web
gunicorn -w 4 -b 0.0.0.0:5000 yolo_webapp:app

2.2.6 配置服务启动

vim /etc/systemd/system/yolo.service

输入如下配置，

[Unit]
Description=YOLO Flask Web App
After=network.target

[Service]
User=python
Group=python
WorkingDirectory=/data/python/cv_flask_web
#ExecStart=/data/python/venv/bin/python /data/python/cv_flask_web/yolo_webapp.py
ExecStart=/data/python/venv/bin/gunicorn -w 4 -b 0.0.0.0:5000 yolo_webapp:app
Restart=always
Environment=PATH=/data/python/venv/bin

[Install]
WantedBy=multi-user.target

服务配置创建后，你需要使用如下命令使配置生效，

systemctl daemon-reloa

然后，你需要使用如下命令启动服务并设置自启动，

systemctl enable yolo.service
systemctl start yolo.service
systemctl status yolo.service

参考文献
==============
https://blog.csdn.net/weixin_45789365/article/details/143312796