This is currently a work-in-progress experiment to replace existing Helm charts for Supabase as they tend to be hard to deploy and to manage and the default Supabase stack - although working great as a single instance or in their SaaS instances - isn't a perfect fit for Kubernetes.
This operator replaces tedious Helm values files with a small set of custom resources that allow an user to quickly deploy a Supabase instance without having to know much (if anything) of the Supabase internals.
- Manage the database instance e.g. making backups, ... that should be done by the Postgres operator or by the user
- Manage your application e.g. run app specific migrations, host your frontend, ...
This operator tries to be as un-opionionated as possible and thereofore does not make assumptions on how you expose APIs to your users (Ingress, GatewayAPI, LoadBalancer service (coming soon))...