Managed and Self-managed for PostgreSQL DBAs
Managed and self-managed covers the choice between running PostgreSQL yourself, on virtual machines or Kubernetes, and using managed services such as Amazon RDS, Aurora, Google Cloud SQL or Neon. Managed services handle patching, backups and failover but limit superuser access and extensions. It is the first cloud stage of the PostgreSQL DBA path.
Stage: Cloud and Operations · Part of the PostgreSQL DBA roadmap · Last reviewed 19 November 2026
What to learn in Managed and Self-managed
- Compare RDS, Aurora, Cloud SQL and Neon on features, control and pricing model
- Explain what a managed service takes over and what remains your responsibility
- Run Postgres on Kubernetes with an operator such as CloudNativePG
- Check parameter groups, extension lists and version support before choosing a service
- Plan connection pooling and read replicas in a managed environment
Managed and Self-managed interview questions
What do you give up when moving from self-managed PostgreSQL to RDS?
Operating system and superuser access, arbitrary extensions and some settings, along with full control of storage layout and upgrade timing. You gain automated backups, patching and failover. Check extension support and parameter limits before committing.
Why use an operator when running PostgreSQL on Kubernetes?
Postgres is stateful, so pods need stable storage, ordered failover, backups and upgrades. An operator such as CloudNativePG encodes this as custom resources, handling replicas, promotion and backup schedules, instead of hand-written StatefulSets and scripts.
Where Managed and Self-managed fits in the PostgreSQL DBA roadmap
Managed and Self-managed belongs to the Cloud and Operations stage (advanced level), which takes about 4 hours. Run it anywhere.
Other units in this stage
Back to the full PostgreSQL DBA roadmap