Panel App is a crowd-sourced repository of information about various gene panels.
Panel App is a project based on Django Framework (v2.1). It uses PostgreSQL as database, AWS SQS as message queue backend and AWS S3 for file storage.
Python version: 3.5
Python dependencies are installed via setup.py.
The previous version of the application used a hosted RabbitMQ instance and the local file system for file storage. This new version has been refactored to work on AWS, leveraging managed AWS services. Using RabbitMQ and the local file system is still possible with the Django settings
./panelapp/panelapp/settings/on-prem.py, but backward compatibility is not fully guaranteed.
All environment are dockerised.
We make a distinction between local development environments and cloud environments, as they use different Dockerfiles and Django settings.
As much as possible, the application follows the Twelve-Factors App design principles.
For more details about local development environment setup and tooling, see ./docker/dev/README.md.
The local environment allows developers to change the code and observe changes.
It also uses LocalStack, to mock AWS SQS and S3 services.
A Docker Compose stack that includes PanelApp (Web and Worker), a PostgreSQL instance and LocalStack is provided.
Web and Worker components share the same codebase but web runs as web-app server while worker runs as Celery.
A Makefile is provided to facilitate dev-time operations.
For details on Cloud environments see ./docker/cloud/README.md.
All environments, except the local-dev environment, are assumed to run on AWS against actual AWS services.
Dockerfiles for cloud are optimised for security and size.
The application is agnostic to the container scheduler platform it runs in (e.g. Kubernetes, ECS).
Docker-compose and Makefile in ./docker/cloud/ are for locally troubleshooting production docker images. They are NOT supposed to be used to deploy the application in any environment.
All contributions are under Apache2 license.