Analytics¶

There only very limited analytics support here. But it’s possible to import your webservers access.log.

Immport your webserver logfile¶

Depends on the format of your logfile. The only supported format at the moment is caddy.

Create a virtualenv on your production server¶

At first, create a analytics user via the django admin interface. I gave it the username ‘analytics’ for convenience. Then create a python virtualenv to run the collect analytics cronjob. It will read the caddy access.log and write the requests into to your django-cast Requests model via a rest-API. You’ll also need pandas in this environment, because the cleanup of request is done in pandas.

apt install virtualenvwrapper  # if you don't already have this installed
apt install libpq-dev  # maybe you'll need that for psycopg2 compilation..
mkvirtualenv -p /usr/bin/python3 your_app_name
pip install pandas

Create a local file with all the required environment variables you need to run django management commands locally - I named it ‘.analytics_env’:

USE_DOCKER=no
DJANGO_AWS_ACCESS_KEY_ID=
DJANGO_AWS_SECRET_ACCESS_KEY=
DJANGO_AWS_STORAGE_BUCKET_NAME=
DJANGO_SETTINGS_MODULE=config.settings.local
USERNAME=analytics
OBTAIN_TOKEN_URL=https://your_domain_name.com/api/api-token-auth/

Be sure that you are now able to run django management commands:

env $(cat .analytics_env | xargs) ./manage.py

Get the api token for your analytics user¶

If you provide the right password for your analytics user, you should now be able to retrieve the api token for that user.

env $(cat .analytics_env | xargs) ./manage.py get_api_token

Don’t forget to add the api token to your ‘.analytics_env’:

API_TOKEN=d387ca7e5d2bf4932f1e9e9c9c4caec808571b39

You’ll need to add two additional environment variables to your ‘.analytics_env’:

REQUEST_API_URL=https://your_domain_name.com/api/request/
ACCESS_LOG_PATH=/var/log/caddy/your_domain_name.access.log

Set up a cronjob to run every hour or so to import your logfile¶

At first, place a shell script named ‘analytics_cron.sh’ in your project dir that you want to execute as a cronjob. It might look like this:

#!/bin/bash

cd $HOME/your_project_dir
(env $(cat .analytics_env | xargs) $HOME/.virtualenvs/your_env_name/bin/python manage.py access_log_import 2>&1) > access_log_import.log

Make this script executable:

chmod +x analytics_cron.sh

And finally create a cronjob running every hour or something like this:

crontab -e
0 * * * * cd $HOME/your_project_dir && ./analytics_cron.sh