2

I'm currently developing an application in Amazon Linux2 using Django, Nginx, Gunicorn, and Postgresql.

I have created and configured a system file for Gunicorn to work when the OS boots, But after starting gunicorn I get the following error...

Jun 08 18:18:04 ip-172-31-16-86.us-east-2.compute.internal systemd[1]: [/etc/systemd/system/gunicorn.service:10] Executable path is not absolute, ignoring: gunicorn --access-logfile - --workers 3 --bind unix:/home/myname/django_project/django_project.sock ......

Jun 08 18:18:04 ip-172-31-16-86.us-east-2.compute.internal systemd[1]: gunicorn.service lacks both ExecStart= and ExecStop= setting. Refusing.

I'm sure the absolute path specified in the file is correct, but I get the error.

How could I solve this problem?

*NOTE: Amazon Linux2 doesn't include www-data groups, so I created using groupadd command.

gunicorn.service

[Unit]
Description=gunicorn daemon
After=network.target
 
 
[Service]
User=myname
Group=www-data
WorkingDirectory=/home/myname/django_project
ExecStart=/home/myname/venv/bin/gunicorn --access-logfile - --workers 3 --bind unix:/home/myname/django_project/django_project.sock django_project.wsgi:application
 
 
[Install]
WantedBy=multi-user.target

Directory structure

home
 │
 └─myname
    │
    ├─django_project
    │    ├─django_project
    │    ├─manage.py
    │    └─django_project.sock
    │  
    └─venv
         └─bin
            └─gunicorn

OS Amazon Linux2

gunicorn 20.1.0

boto 2.49.0

Django 3.2.4

django-ses 2.0.0

Tio
  • 944
  • 3
  • 15
  • 35

1 Answers1

0

Hopefully others will be have better answers or suggestions but this discussion emphasizes to double check the existence of the path to the Gunicorn executable or making sure your environment recognizes that.

This discussion also further emphasizes checking everything is installed correctly

Viet Than
  • 164
  • 1
  • 10