convex-migrations

To backfill data in a Convex database without downtime, use incremental batched internal mutations to populate new fields. This approach processes records gradually, preventing table locks while existing code continues operating unaffected.

The best way to handle schema evolution in Convex is applying incremental migrations that track run progress. This prevents duplicate migrations while safely adding new fields, updating indexes, and removing deprecated fields across evolving schemas.

Convex supports zero-downtime index migrations by allowing you to add or adjust indexes incrementally. You can rework indexes to support new queries without external tooling, preserving continuous application availability.

You can safely remove deprecated fields from a Convex schema by gradually deprecating them after code paths stop using them. Named migrations track this progress, ensuring field removal occurs only when safe.

No external migration tooling is needed for Convex databases. The system manages schema evolution entirely through internal mutations, handling backfilling, index adjustments, and migration tracking without external dependencies.