CONCEPT Cited by 1 source

Additive schema change

Definition

An additive schema change is a schema modification that only adds new information without removing, renaming, or reinterpreting existing fields. The canonical additive changes are:

• Add column (nullable, no default required for existing records)
• Widen numeric precision (e.g., INT → BIGINT, where semantic meaning is unchanged)

Additive changes preserve backward compatibility: older consumers that don't know about the new field continue functioning (they see null or ignore the extra column). No historical data reinterpretation is required.

Why systems restrict to additive-only

In a distributed CDC pipeline, non-additive changes (type narrowing, column rename, primary key change) require coordination across multiple independent systems with different deployment cadences. Restricting automated evolution to additive changes:

1. Preserves backward compatibility
2. Avoids historical replay complexity
3. Minimises risk of breaking existing consumers
4. Makes automation tractable (no ambiguity resolution needed)

Non-additive changes are still possible but require manual migration, re-onboarding, or coordinated deployment.

Pinterest's framing

"This is a deliberate tradeoff. Additive-only changes preserve backward compatibility, avoid historical replay complexity, and minimize the risk of breaking existing consumers." (Source: sources/2026-06-24-pinterest-automated-schema-evolution-in-pinterests-next-generation-db)

Seen in

• sources/2026-06-24-pinterest-automated-schema-evolution-in-pinterests-next-generation-db — canonical instance: Pinterest restricts automated schema evolution to additive changes across their CDC pipeline