Skip to content

Latest commit

 

History

History
31 lines (24 loc) · 1.44 KB

deploying.md

File metadata and controls

31 lines (24 loc) · 1.44 KB
Raw

Deploying

Deploying WTFIX to staging or production environments generally involves the following steps:

  • Set the WTFIX_SETTINGS_MODULE environment variable to config.settings.staging or config.settings.production as the case may be.

  • Ensure that the DEBUG configuration parameter is set to False.

  • If the pipeline has been configured to include the RESTfulServiceApp, then you will also need to configure a WSGI HTTP server for hosting the APIs (Flask's built-in server is not suitable for production). WTFIX comes with support for gunicorn out of the box, though any well-supported WSGI container should do.

    gunicorn --workers 1 --bind unix:<path_to_your_project>/wtfix.sock 'config.wsgi:get_wsgi_application(session_name="default")'

NOTE: many FIX servers do not allow multiple connections using the same logon credentials, so it probably noes not make sense to run more than one worker process. You should also consider monitoring the above process with something like supervisord:

[program:wtfix]
command=<path_to_your_gunicorn_binary>/gunicorn --workers 1 --bind unix:<path to your project>/wtfix.sock 'config.wsgi:get_wsgi_application(session_name="default")'
directory=<path_to_your_project_dir>
autostart=true
autorestart=true
stderr_logfile=/var/log/wtfix.err.log
stdout_logfile=/var/log/wtfix.out.log