Flask - Deployment with Gunicorn, Nginx and Supervisor, Supervisor error log

Question

I've deployed a Flask app with Gunicorn + Nginx + Supervisor. It doesn't works. Searching online turned up several reports of similar issues, but none with an explanation that fit our circumstances or a fix that resolved the issue. Here said to turn off daemonize gunicorn. If I'm right, it's not my case.

It seems to be that something start a service that listening 8000 port already. Supervisor sends errors to the log file every second.

What can help me?

The error from the Supervisor app-stderr.log:

[2015-02-19 18:56:19 +0300] [964] [INFO] Starting gunicorn 19.2.1
[2015-02-19 18:56:19 +0300] [964] [INFO] Listening at: http://127.0.0.1:8000 (964)
[2015-02-19 18:56:19 +0300] [964] [INFO] Using worker: sync
[2015-02-19 18:56:19 +0300] [1078] [INFO] Booting worker with pid: 1078
 * Running on http://127.0.0.1:5000/ (Press CTRL+C to quit)
 * Restarting with stat
[2015-02-19 18:56:20 +0300] [1081] [INFO] Starting gunicorn 19.2.1
[2015-02-19 18:56:20 +0300] [1081] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:20 +0300] [1081] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:21 +0300] [1081] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:21 +0300] [1081] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:22 +0300] [1081] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:22 +0300] [1081] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:23 +0300] [1081] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:23 +0300] [1081] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:24 +0300] [1081] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:24 +0300] [1081] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:25 +0300] [1081] [ERROR] Can't connect to ('localhost', 8000)
[2015-02-19 18:56:25 +0300] [1078] [INFO] Worker exiting (pid: 1078)
[2015-02-19 18:56:25 +0300] [1122] [INFO] Booting worker with pid: 1122
 * Running on http://127.0.0.1:5000/ (Press CTRL+C to quit)
 * Restarting with stat
[2015-02-19 18:56:27 +0300] [1148] [INFO] Starting gunicorn 19.2.1
[2015-02-19 18:56:27 +0300] [1148] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:27 +0300] [1148] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:28 +0300] [1148] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:28 +0300] [1148] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:29 +0300] [1148] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:29 +0300] [1148] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:30 +0300] [1148] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:30 +0300] [1148] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:31 +0300] [1148] [ERROR] Connection in use: ('localhost', 8000)
[2015-02-19 18:56:31 +0300] [1148] [ERROR] Retrying in 1 second.
[2015-02-19 18:56:32 +0300] [1148] [ERROR] Can't connect to ('localhost', 8000)
[2015-02-19 18:56:32 +0300] [1122] [INFO] Worker exiting (pid: 1122)
[2015-02-19 18:56:32 +0300] [1206] [INFO] Booting worker with pid: 1206
 * Running on http://127.0.0.1:5000/ (Press CTRL+C to quit)
 * Restarting with stat
[2015-02-19 18:56:32 +0300] [1211] [INFO] Starting gunicorn 19.2.1
[2015-02-19 18:56:32 +0300] [1211] [ERROR] Connection in use: ('localhost', 8000)

My supervisor.conf:

[program:app]
command = /home/www/app/flask/bin/gunicorn app:app -b localhost:8000 --preload
directory = /home/www/app
user = webhost

Netstat -tulpin

Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name
tcp        0      0 0.0.0.0:80              0.0.0.0:*               LISTEN      930/nginx
tcp        0      0 0.0.0.0:22              0.0.0.0:*               LISTEN      855/sshd
tcp        0      0 127.0.0.1:8000          0.0.0.0:*               LISTEN      1392/python3
tcp6       0      0 :::22                   :::*                    LISTEN      855/sshd

And ps aux | grep python

root       954  0.1  2.4  60564 12440 ?        Ss   18:56   0:01 /usr/bin/python /usr/bin/supervisord -c /etc/supervisor/supervisord.conf
webhost   1392  0.0  3.2  67148 16512 ?        S    19:00   0:00 /home/www/app/flask/bin/python3 /home/www/app/flask/bin/gunicorn app:app -b localhost:8000
webhost   2124  5.2  4.2  86072 21344 ?        S    19:12   0:00 /home/www/app/flask/bin/python3 /home/www/app/flask/bin/gunicorn app:app -b localhost:8000
webhost   2126  5.2  3.2  67148 16488 ?        S    19:12   0:00 /home/www/app/flask/bin/python3 /home/www/app/flask/bin/gunicorn app:app -b localhost:8000
webhost   2130  0.0  0.1  11744   920 pts/0    S+   19:13   0:00 grep --color=auto python

UPD. supervisor config is above

nginx.conf

server {
    location / {
        proxy_pass http://127.0.0.1:8000;
    }
    location /static {
        alias  /home/www/myapp/app/static/;
    }
}

File sctructure

myapp
  |
  |-app-
  |    |
  |    __init__.py  
  |
  run.py
  |
  config.py

__init__.py

from flask import Flask

app = Flask(__name__)
app.config.from_object('config')

from app import views

run.py

#!flask/bin/python
from app import app
app.run()

config.py

# For valid forms
CSRF_ENABLED = True
SECRET_KEY = 'never-guess'

Answer 1

I had the same issue and by adapting Alwx's answer my solution was to only run the app if the module is called as main (run.py):

from my_app import create_app

app = create_app()
if __name__ == '__main__':
    app.run(host='0.0.0.0', port=8080, debug=True)

Then I run it with gunicorn:

gunicorn run:app

And I can still run it as before with just python run.py

Answer 2

I fugured out. Thanks to @Ibrahim I started an app via gunicorn in run.py

#!flask/bin/python
from app import app
app.run()

So gunicorn created process and app.run() too. And they conflict each other. Just deleted that string.