Connecting a data source

This guide covers choosing and configuring a source connector — the first step of any new pipeline.

Choosing a connector type

You haveUse
A relational database you can grant replication access toPostgres
Files landing in object storageAmazon S3
An existing event streamKafka
A third party that can push to a URL you controlWebhook

Authenticating

Each connector type has its own credential shape — see the Connectors endpoint for the full parameter list per type.

Least-privilege credentials

Grant the connector only the access it needs — read-only replication for Postgres sources, a scoped IAM role for S3, never a superuser or admin credential.

Verifying the connection

A newly created connector reports health: degraded until its first successful read or write, then flips to healthy. If it goes straight to down, double check credentials before opening a support ticket.

Next steps

With a source connected, follow Building your first pipeline to wire it into a full pipeline.