Supabase Pipelines: GitHub Discussion

Introducing Supabase Pipelines.

Pipelines replicates your Postgres changes to an analytics destination in near real time, so heavy analytics queries run there instead of competing with your app on the primary database. It's powered by Supabase ETL, our open-source engine, and you set it up in the Dashboard: pick your tables, pick a destination, and data starts flowing after an initial sync.

BigQuery is the first destination. ClickHouse, Snowflake, and DuckLake are in progress, and there's a request form if you want to register interest in one of those.

A few details:

• It captures inserts, updates, deletes, and truncates through Postgres logical replication
• Supported schema changes (add, rename, remove columns) replicate to the destination automatically
• It's available on all paid plans, no access request
• Pricing is usage-based and subject to change during the alpha

It's still alpha, so we're actively improving performance and adding destinations. If you're using it for analytics, reporting, ML, or archiving, we'd like to hear how it's going and which destinations you want next.

Docs: https://supabase.com/docs/guides/database/replication/pipelines Blog: https://supabase.com/blog/supabase-pipelines-public-alpha Open-source engine: https://github.com/supabase/etl