Supabase (self-hosted, core)

Why deploy on Railyard

Railyard provisions all 7 services in this stack together — supabase-db, supabase-auth, supabase-rest, supabase-realtime, supabase-storage, supabase-meta, supabase-studio — wired to reach each other over the private network, on a server you already pay for.

Deploying Supabase (self-hosted, core) is one piece of a bigger stack you can keep building — add a database, a worker, a cache, or another app, all on the same servers you already pay for.

Good for

• A self-hosted Postgres + Auth + Storage + Realtime backend
• A Firebase-style backend for a team that wants to own the database

Good to know

• Every service gets its own Railyard URL instead of sitting behind one API gateway — Supabase's own compose routes everything through Kong/Envoy, which is skipped here the way nginx is skipped elsewhere in this catalog
• This runs as a single project, same as upstream self-hosted Supabase — Studio doesn't support multiple organizations or projects the way Supabase Cloud does
• Known gap: supabase-realtime still won't boot cleanly — its migrator writes an unqualified public.schema_migrations, which collides with the bookkeeping table the supabase/postgres image's own bootstrap already creates there. Upstream avoids this by mounting an extra init script that pre-creates a dedicated _realtime schema before Realtime ever starts; this template has no equivalent bootstrap step yet, so redirecting Realtime's search_path there fails too (the schema doesn't exist). Needs a Railyard-side way to run one-time setup SQL against supabase-db before dependent services boot