git clone https://github.com/SeanEstey/flask_celery_skeleton --branch <b_name>
cd flask_celery_skeleton

Follow instructions in requirements/pkg_list.txt and requirements/requirements.txt

python setup.py

This will copy nginx virtual host file and setup logrotate.d

Create "db_auth.py" in Bravo root directory:

user = "db_user"
password = "db_pw"

Start RabbitMQ daemon:

$ rabbitmqctl start_app

Run app:

python run.py

Arguments

-Start with celerybeat:

-c, --celerybeat

-Start in debug mode:

-d, --debug

If running in foreground, kill with CTRL+C. This will kill Celery workers.

If running in background, get pid:

$ps aux | grep -m 1 'python main.py' | awk '{print $2}'

Now kill it using that PID:

$kill -9 <PID>

(May need to run twice)

Monitor Celery worker(s) with Flower:

pip install flower

To run it:

flower --url_prefix=flower --basic_auth=user1:password1

To access it remotely through the browser, add the following to the nginx virtual_host file:

server {
    listen 80;
    server_name ip_or_hostname.ca;

    location /flower/ {
        rewrite ^/flower/(.*)$ /$1 break;
        proxy_pass http://127.0.0.1:5555;
        proxy_set_header Host $host;
    }

Restart nginx:

service nginx restart

It should now be accessible and secured via http://ip_or_hostname.ca/flower

To free memory not released by abberant python/celery processes:

$ sync && echo 3 | sudo tee /proc/sys/vm/drop_caches