convex-migration-helper

Zero-downtime Convex schema migrations use a widen-migrate-narrow workflow, expanding schemas to handle dual read/write formats, backfilling data, then narrowing schemas after verifying completion to prevent deploy failures.

Breaking Convex schema changes are handled using online migration patterns supporting dual read/write formats, enabling applications to remain online while batched data backfills complete safely.

You need the widen-migrate-narrow workflow for Convex when applying breaking changes like adding required fields, changing types, renaming or removing fields, or splitting and merging tables.

Data backfill during online Convex migrations is executed using @convex-dev/migrations for batched resumable processing, featuring dry runs and status monitoring to verify completion before narrowing schemas.

Convex supports online migrations for nested-to-relational data by enforcing a widen-migrate-narrow workflow that handles dual read/write formats during the migration window without application downtime.

Convex deployments fail when changing schema types because new schemas become incompatible with data at rest, requiring online migration workflows to handle dual formats before finalizing type changes.