Hey folks, looking for some help and inputs to implement multi tenancy in our implementation of Phoenix. Is there support for authorization at the project level or anyone other methods to implement secure multi-tenancy aside from using a project per tenant?
Hey Akshay, one of the developers of Phoenix Support here. Thanks for using Phoenix! I think we do at some point plan on supporting multi-tenancy within a single Phoenix but this is currently not how we think of Phoenix - Phoenix is basically the right size for a single tenant and the users that have access to it should be explicitly granted permissions to that tenant. You can deploy multiple phoenix instances talking to a single DB by giving them two different schemas. You can then setup a reverse proxy in front so you cat /tenant/a tenant/b. For enterprise constructs like spaces and organizations you might also consider Arize Ax. I'll try to get something up in our documentation. I'll let Priyan J. and Srilakshmi C. take it from here.
Thanks Mikyo! That makes sense. Priyan J. , Srilakshmi C. . Any other pointers to keep in mind while implementing the above architecture? Is there any recommended upper limit for the number of tenants/schemas for Postgres?
