PlanetScale

Recently shipped PlanetScale Postgres and Neki databases on Microsoft Azure, plus retryable and aggressive-cutover deploy requests.

Azure joins AWS and Google Cloud. Postgres and Neki databases can now be created in Azure regions with the same features as other clouds — high availability, branching, Query Insights, automated backups, and Metal clusters. Vitess databases aren't available on Azure yet.

Deploy requests recover instead of restarting. A partially failed Vitess migration can retry only the failed tables with pscale deploy-request retry, leaving healthy tables at their existing progress. Aggressive cutover is now opt-in per deploy request from the CLI and API, changeable any time before you deploy, and flagged in show and queue.

MCP clients can read branch metrics. planetscale_get_branch_metrics pulls CPU, memory, disk, replication lag, and connections into an AI assistant, for Vitess, Postgres, and Neki. The default summary covers the last hour.

Parameter changes now say what they need first. CLI v0.346.0 adds a Requires update column for Postgres branch and Neki parameters, plus a requires_update JSON field. Terraform provider v1.13.0 manages Neki router parameters and filters databases by kind; if you manage Postgres bouncers with Terraform, upgrade, since earlier providers can't read them after recent changes to how PgBouncer parameters are reported.

Still standing from earlier releases:

• Postgres dedicated read replicas scale independently of the primary cluster.
• MySQL databases import with no downtime through pscale branch vtctl move-tables, which replaces the deprecated pscale workflow.
• pscale keyspace parameters views and sets VTTablet and MySQL parameters per keyspace.
• TIN 1.0.4 added word stemming and bounded top-k score tie-breaking.
• database-js 2.0.0 dropped parameter interpolation and requires Node 22.