Skip to main content
Version: v4.2.1

Migrations

This page covers migration behavior for both database schema and storage data in OpenReader.

Startup migration behavior​

By default, the shared entrypoint runs migrations automatically before app startup in:

  • Docker container startup
  • pnpm dev
  • pnpm start

Startup migration phases:

  • DB schema migrations (pnpm migrate)
  • Storage/data migration (pnpm migrate-fs) for legacy filesystem content into S3 + DB rows
info

In most setups, you do not need to run migration commands manually because startup handles this automatically.

Schema history​

Migrations are applied in order. All of the following ship in v3.0.0; an instance upgrading from v2.2.0 applies 0001–0004 in a single startup pass.

MigrationDialectsWhat it does
0001_tts_segmentsSQLite + PostgresCreates the original single-table tts_segments used by server-side TTS segment caching.
0002_add_segment_key_to_tts_segmentsSQLite + PostgresAdds the segment_key column to tts_segments for stable locator-independent segment identity.
0003_tts_segments_v2_splitSQLite + PostgresReplaces tts_segments with a normalized two-table model: tts_segment_entries (one row per document segment + locator identity) and tts_segment_variants (one row per settings combination, holding the cached audio key, status, and alignment). Drops the original tts_segments table — no released build (v2.2.0 or earlier) ever populated it, so there is no production data to migrate.
0004_admin_panelSQLite + PostgresCreates admin_providers (encrypted shared TTS provider rows) and admin_settings (runtime site-feature config), and adds the is_admin column to the user table. Backs the Admin Panel.

To skip automatic startup migrations:

  • Set RUN_DRIZZLE_MIGRATIONS=false
  • Set RUN_FS_MIGRATIONS=false
warning

If you disable startup migrations, ensure your deployment process runs migrations before serving traffic.

Apply migrations​

In most cases, you do not need manual migration commands because startup runs migrations automatically.

pnpm migrate applies migrations for one database target:

  • Postgres when POSTGRES_URL is set
  • SQLite when POSTGRES_URL is unset

You can always override the target explicitly with --config.

# Run pending migrations for one target:
# - Postgres if POSTGRES_URL is set
# - SQLite if POSTGRES_URL is unset
pnpm migrate

# Run storage migration (filesystem -> S3 + DB)
pnpm migrate-fs

# Dry-run storage migration without uploading/deleting
pnpm migrate-fs:dry-run

Generate migrations​

pnpm generate is a two-phase script for contributors and schema changes:

  1. Better Auth schema generation — runs the Better Auth CLI twice (once for SQLite, once for Postgres) to produce auto-generated Drizzle schema files for auth tables (user, session, account, verification).
  2. Drizzle migration generation — runs drizzle-kit generate for both drizzle.config.sqlite.ts and drizzle.config.pg.ts, producing SQL migration files from all schema files (app + auth).
note

Most users do not need to run pnpm generate. Use it when contributing or when you have changed Drizzle schema files and need new migration files.

Schema ownership​

Auth tables are owned by Better Auth. Their Drizzle schema definitions are auto-generated and should not be hand-edited:

  • src/db/schema_auth_sqlite.ts
  • src/db/schema_auth_postgres.ts

App-specific tables are manually maintained in the standard Drizzle schema files:

  • src/db/schema_sqlite.ts
  • src/db/schema_postgres.ts

Both sets of schema files are included in the Drizzle configs, so drizzle-kit generate and drizzle-kit migrate handle all tables together.

When app schema changes (for example tts_segments), keep these in sync:

  • src/db/schema_sqlite.ts
  • src/db/schema_postgres.ts
  • drizzle/sqlite/*.sql + drizzle/sqlite/meta/_journal.json
  • drizzle/postgres/*.sql + drizzle/postgres/meta/_journal.json
# Full pipeline: Better Auth CLI + Drizzle generate (both dialects)
pnpm generate