Because of the increasing traffic on this site, I also use Gunicorn as the middleware between Nginx and Flask, otherwise the test server that comes with Flask is really unable to withstand such a large traffic .

# Start Gunicorn

There is no need to talk too much about startup. After configuring the conf file, use the following command to start.

gunicorn -c gunicorn.py app:app

# Get the Gunicorn process ID

Use pstree to query the process tree.

pstree -ap|grep gunicorn

The return should look like this:

root@web-dev:~/# pstree -ap|grep gunicorn
   |-gunicorn,3970631 /usr/bin/gunicorn -c gunicorn.py app:app
   | |-gunicorn,3971828 /usr/bin/gunicorn -c gunicorn.py app:app
   | |-gunicorn,3971829 /usr/bin/gunicorn -c gunicorn.py app:app
   | |-gunicorn,3971830 /usr/bin/gunicorn -c gunicorn.py app:app
   | `-gunicorn,3971831 /usr/bin/gunicorn -c gunicorn.py app:app
   | | |-grep,3972703 --color=auto gunicorn

3970631 is Gunicorn's process ID.

# Restart Gunicorn

Use kill -HUP to bring the process ID to restart the process.

kill -HUP 3970631

After executing the above command, execute "pstree -ap|grep gunicorn" again, we can easily find that, except the main process, other Gunicorn processes have been destroyed, and a new process has been created (the process ID has changed).

# Quit Gunicorn

kill -9 3970631