PlanetScale Branch Manager

PlanetScale branches are isolated copies of your database schema, similar to Git branches for code. You can create a feature branch, safely test schema changes (ALTER TABLE, CREATE INDEX), then open a deploy request to merge those changes into production — without locking tables or causing downtime.

A deploy request is PlanetScale's equivalent of a pull request for database schema changes. It shows the DDL diff between your feature branch and the target (usually main/production), lets your team review the changes, and deploys them using Online Schema Change (OSC) — zero-downtime migrations that don't lock tables.

PlanetScale uses gh-ost (GitHub Online Schema Change) under the hood. Instead of ALTER TABLE which locks the table, it creates a shadow copy of the table, applies changes, syncs data via triggers, then swaps the tables atomically. Reads and writes continue uninterrupted throughout the migration.

Yes. Connect to your branch via pscale connect or pscale shell, then run DDL statements. However, note that PlanetScale's foreign key constraint support has limitations — they recommend using application-level foreign key enforcement instead of database-level constraints for the online migration system to work correctly.

pscale connect opens a local MySQL-compatible proxy on a port (default 3306) — use this to connect your app or any MySQL client. pscale shell opens an interactive MySQL REPL directly in your terminal — useful for running quick queries, schema commands, or seeding data on a specific branch.

For Prisma: use the @planetscale/database driver with DATA_PROXY or direct connection. Set databaseUrl in schema.prisma. For Drizzle: use drizzle-orm/planetscale-serverless with @planetscale/database for edge/serverless environments, or drizzle-orm/mysql2 for Node.js servers. Both support the PlanetScale branching workflow.