Deployment of flask project on VPS

(Updated: )

Intro

This article presents my experience on how to deploy a Flask web application on my VPS(Virtual Private Server).

I designed an OCR-based application for paper-document recognition and retrieval, which was written in Python. For ease to use, I made it a simple web app and deployed it on my rented VPS.

Tips:

FLASK: Flask is a micro web framework written in Python.

VPS: A virtual private server (VPS) is a virtual machine sold as a service by an Internet hosting service.

from en.wikipedia.org

VPS setup

requirements:

  • remote control the VPS by ssh
  • have a user with root access

See my other article for details.

general setup

setup firewall

UFW, full name Uncomplicated Firewall, is a Firewall tool implemented through iptables.UFW is a encapsulation of iptables, whose rules are too complex for beginners.This article is also an extension on how to enhance VPS security. After you modify the SSH port and log in the user, it is a good idea to know which ports you have access to on your VPS and manage them with UFW to prevent abuse.Linux 下的防火墙 ufw

here is the command I use to set up the firewall

sudo apt install -y ufw

sudo ufw default allow outgoing

sudo ufw default deny incoming

sudo ufw allow ssh # make sure you can use ssh for remote access

sudo ufw allow 5000 # for test

sudo ufw enable

sudo ufw status
sudo ufw status

setup Flask Project

This is the file tree of the project for configuration reference.

1
2
3
4
5
6
7
8
9
10
11
12
13
➜  ~ pwd
/home/tom
➜  ~ tree -L 3
.
└── Projects
    └── DocumentOCR
        ├── app
        ├── config.py
        ├── doc
        ├── __pycache__
        ├── requirements.txt
        ├── run.py
        └── venv

Tip: prepare a requirement.txt
Include all the package need to be installed in this flask project. if does not know what are those packages, can use command pip freeze

then use git clone to get the project files from your github repository

PS: you can use scp as well.

Install environment requirements

sudo apt install -y python3-pip

sudo apt install -y python3-venv

git clone repository(my example)

mkdir Projects

cd Projects

git clone https://github.com/GoldExperience/DocumentOCR.git

Create and active Virtual environment

python3 -m venv {Your Project Dir}/venv

source venv/bin/activate

Then we can install all requirements with requirements.txt

pip install -r requirements.txt

if this step has problems, you can install the requirements manually using pip install {package name} as well.

setup nginx & gunicorn

Because the poor performance of wsgi of flask, we need use nginx + gunicorn to get better performance and security.

Nginx is a web server that can also be used as a reverse proxy, load balancer, mail proxy and HTTP cache.

and gunicorn means Multi-process/multi-threading/automatic-restart

sudo apt install -y nginx

sudo rm /etc/nginx/sites-enabled/default

sudo vim /etc/nginx/sites-enabled/$ProjectName

1
2
3
4
5
6
7
8
9
10
11
12
13
14
server{
        listen 80;
        server_name {your ip address};

        location /static{
                alias /home/tom/Projects/{ProjectName}/app/static;
        }

        location /{
                proxy_pass http://localhost:8000;
                include /etc/nginx/proxy_params;
                proxy_redirect off;
        }
}

the /home/tom/Projects/{ProjectName}/app/static is the location of the static files of your Flask Project.

if you cannot save in vim, try this:

:w !sudo tee % > /dev/null

setup ufw again

sudo ufw allow http/tcp

sudo ufw delete allow 5000

sudo ufw enable

sudo systemctl restart nginx

gunicorn start

pip install gunicorn

gunicorn -w {worker} run:app

worker = core * 2 +1

using supervisor manage gunicorn

In order for the system to continue running even after the remote connection is closed, you need to use the Supervisor to manage the gunicorn.

sudo apt-get install -y supervisor

then setup config file for supervisor

sudo vim /etc/supervisor/conf.d/$ProjectName.conf

1
2
3
4
5
6
7
8
9
10
[program:{ProjectName}]
directory=/home/tom/Projects/{ProjectName}
command=/home/tom/Projects/{ProjectName}/venv/bin/gunicorn -w {worker} run:app
user={UserName}
autostart=true
autorestart=true
stopasgroup=true
killasgroup=true
stderr_logfile=/var/log/{ProjectName}/{ProjectName}.err.log
stdout_logfile=/var/log/{ProjectName}/{ProjectName}.out.log

create log file dir

sudo mkdir -p /var/log/$ProjectName

sudo touch /var/log/$ProjectName/$ProjectName.err.log

sudo touch /var/log/$ProjectName/$ProjectName.out.log

restart supervisor

sudo supervisorctl reload

Change

sudo vim /etc/nginx/nginx.conf
add this within block http
client_max_body_size {upt to you}M;

the restart nginx

sudo systemctl restart nginx

conclusion

files need to be setup about specific project

  • /etc/nginx/sites-enabled/$ProjectName

  • /etc/supervisor/conf.d/$ProjectName.conf

Other Tips

Tips: for opencv error:ImportError: libGL.so.1: cannot open shared object file: No such file or directory
try: sudo apt install libgl1-mesa-glx

test_before nginx:

export FLASK_APP=run.py

flask run –host=0.0.0.0