-
Notifications
You must be signed in to change notification settings - Fork 0
Home
- Python v3.6 (& pip)
- Node.js v10 (& npm)
- openctm
The phenome10k web application is written in Python3 as a Flask application. Many of the views are written using Vue.js as .vue files, and require Node both for transpilation to JavaScript and for on-demand server-side rendering.
As part of the upload process, 3D models are converted to CTM files for viewing in the browser using the JSC3D library. This process uses the ctmconv tool from the openctm library.
By default the application stores data in the working directory, including using sqlite as the database. The application can be configured using the following environment variables:
| Key | Description |
|---|---|
| SECRET_KEY | Flask encryption key |
| DATABASE_URL | URL of database to connect to, including protocol |
| THUMB_DIRECTORY | Path of directory to store automatically generated thumbnails to |
| MODEL_DIRECTORY | Path of directry to save original uploaded models to |
| UPLOAD_DIRECTORY | Path of directory to save all other user files to |
| LOG_FILE | Filepath of where to write logs to |
| MAIL_DEFAULT_SENDER | From address of emails sent by the system |
| ADMIN_EMAIL | Email of admin user to send admin-related requests to |
These may also be added to a .env file either in the working directory or the parent directory (parent takes priority).
The production environment comprises two load balancers in HA-pair configuration, two app servers, and one data server for mysql database and nfs file storage. The storage server is to be replaced in future by separate database and nfs clusters.
The private variables used on production are stored in the museum's Vault service.
The application can be deployed to production using the ansible script included in the project's git repository. Instructions on how to run the script are included in the README.md file. There is also a vagrant-based local test/development available.