Skippr Docs

Appearance

When to Choose Redshift for CDC Final State ​

May 2026

Choose Redshift for CDC final state when you want final tables in Redshift and are comfortable with the documented S3 staging and MERGE-based apply path.

Short Answer ​

Choose Redshift when Redshift is where you want the final-state tables to live and the documented S3 staging plus staging-table MERGE workflow fits your environment. Skippr loads through a staging bucket, applies guarded merges, and creates the same automatic CDC helpers described in the destination docs: _skippr_order_token columns and tombstone tables.

That makes Redshift a good fit when you already want Redshift to own the reconciled warehouse table and you can provide the required AWS and Redshift setup. The decision is less about whether Redshift can hold rows and more about whether the COPY-and-MERGE path is the warehouse contract you want to run every day.

Why Teams Struggle with This ​

Redshift CDC has more visible infrastructure than a direct-write destination. If the team does not want to think about a staging bucket, IAM role, and Redshift write permissions, that should influence the choice before implementation starts.

  • The documented load path depends on a staging S3 bucket and an IAM role that Redshift can use to read from it.
  • Final-state reconciliation still depends on business keys and order-token comparison, not on COPY alone.
  • Deletes still need tombstone protection so an older replay cannot bring a deleted row back.
  • The Redshift database user needs write access to the target schema in addition to the AWS-side setup.

How Skippr Handles It ​

Skippr makes this choice easier because Redshift is not presented as a vague sink. The docs are explicit that the destination uses staging-table MERGE semantics and creates the order-token and tombstone structures automatically.

That means you can choose Redshift for what it actually does in the CDC path: stage, merge, protect against stale writes, and preserve delete intent in a way the guarantees page already explains.

  • Redshift CDC is documented as exactly-once final-state reconciliation.
  • The load path is explicit: S3 staging plus Redshift COPY workflow and MERGE apply.
  • Order-token columns are created automatically on CDC-managed tables.
  • Tombstone tables are created automatically to prevent ghost resurrections.

What the First Useful Version Looks Like ​

Choose Redshift when you want Redshift to be the place where CDC final state is reconciled and the documented S3 staging model is already acceptable for your environment.

If that operational shape fits, Redshift gives you a clear warehouse-side contract instead of a generic landing zone.