PlanetScale vs Apache Airflow

The Postgres line carries the new capability: vectorscale extension support, storage configurable at database creation, out-of-memory alerts by both email and webhook, a customisable log line prefix, a Discovery Tool for evaluating existing databases, and updated pricing for private connection traffic. Database Traffic Control introduced resource budgets and has since gained warning thresholds and CLI management. The surface work is constant — API endpoints for backup policies, maintenance schedules, schema recommendations, deploy-request storage checks and invoice status, a Terraform provider at v1, and an Insights-only MCP server.

PlanetScale is doing two things at once: closing the gap between its Postgres offering and its MySQL heritage, and making every operation scriptable. The API, Terraform and MCP work share a single goal — nothing should require the dashboard — and the schema-recommendations feature landing simultaneously as a REST endpoint and an MCP tool is the clearest statement of it. Traffic Control is the exception, launched as a named product surface and then immediately given the same CLI and threshold treatment.

Expect the remaining Postgres parity items and more MCP tools mirroring existing REST endpoints, following the schema-recommendations pattern where both arrived together.