Autoscaling and Scale-to-Zero

Makes Lakebase connections production-grade. Following the lecture-demo pattern, learners compare OAuth tokens and native passwords, handle the 60-minute token lifetime with refresh, and build a token-refreshing connection pool in both SQLAlchemy and raw psycopg — sized against the endpoint's connection ceiling and observed through pg_stat_activity. The course concludes with the Lakebase Data API, showing how the authenticator role brokers identity via SET ROLE to enforce the same privilege model over HTTP.

Turns copy-on-write branching into everyday practice. Following the lecture-demo pattern, learners create branches four ways, recover data with point-in-time branches, manage the branch lifecycle (protect, reset, delete), and choose a branching strategy with per-workload TTLs — including branch-per-PR CI/CD. Branches are the feature that lets developers pair a code branch with a database branch. Two real OrderOps admin apps, one bound to each branch, make isolation tangible: learners edit orders in each app and prove changes never cross branches, then graduate a schema change from the dev branch to production by re-running the migration and shipping the new app version.

Comprehensive auditing, monitoring, and observability using system tables, query history analysis, and proactive alerting for DBSQL environments.

Note: Databricks Academy is transitioning to a notebook-based format for classroom sessions within the Databricks environment, discontinuing the use of slide decks for lectures. You can access the lecture notebooks in the Vocareum lab environment.