convex-migration-helper

To run a zero-downtime Convex schema migration, you apply a three-step workflow: widen the schema, backfill data, then narrow the schema. This process uses batched migrations to deploy breaking changes safely without downtime.

The best way to backfill data during a Convex migration is using the @convex-dev/migrations component. It provides batched data processing, handles nested-to-relational mappings, and preserves data integrity during online updates.

Yes, you can dry-run a data migration in Convex before execution. The @convex-dev/migrations component supports dry-runs and status monitoring, allowing you to validate schema changes safely before full deployment.

You need to widen the schema for a Convex deployment when applying breaking changes that require backfilling data. Widening allows new and old application versions to read data simultaneously during a zero-downtime migration.

Yes, the @convex-dev/migrations component supports batched migrations. It enables safe online data processing by breaking large migration tasks into smaller batches, ensuring minimal downtime and safe schema transitions.