Architecture

High-level architecture

High-level architecture of a Django application

A SERIMA instance is made of the following components:

  • Web application: the Django project (governanceplatform), served by Gunicorn or Apache mod_wsgi. Static files are served by the application itself, with WhiteNoise.

  • Theme: the templates, styles and translations of the interface, in a separate Git repository checked out in the theme folder (see Installation).

  • PostgreSQL: all persistent data.

  • Redis: the message broker between the web application and the background workers.

  • Celery worker: runs the slow tasks outside the web requests — report generation (DOCX, PDF and the charts rendered with Kaleido), security objectives exports and imports.

  • Celery beat: triggers the scheduled tasks, such as reminders and data retention (see Installation).

  • File storage: generated reports and import/export files, written to PATH_FOR_REPORTING_PDF.

  • Email server: notifications, reminders and password resets.

Models

The overview shows the models of the four applications and their relations, without fields. The diagrams below show each application in detail; models from other applications appear as plain boxes. The django-parler translation models (*Translation) are left out: a translatable model is marked <TranslatableModel> in its header.

Overview of the models of the governance, incidents, security objectives and reporting applications

Governance

Models of the governance application: users, companies, regulators, sectors and regulations

Incident notification

Models of the incidents application: incidents, workflows, questions and impacts

Security objectives

Models of the security objectives application: standards, domains, objectives, measures and answers

Reporting

Models of the reporting application: projects, risk data, observations, templates and generated reports

The diagrams are generated with make models.