Postgres
PostgreSQL (Postgres) is the world’s most advanced open source relational database. It is ACID-compliant, extensible, and runs everywhere from laptops to managed cloud instances. We adopt it as the default OLTP and analytics-adjacent store when a relational model fits.
Blurb
PostgreSQL is a powerful, open source object-relational database system with over 35 years of active development that has earned it a strong reputation for reliability, feature robustness, and performance.
Summary
What it is: Object-relational engine with SQL, JSONB, extensions (PostGIS, pgvector, and many others), logical replication, and permissive licensing. Same project whether self-hosted, on RDS/Aurora/AlloyDB, or embedded.
When to use: New application databases, warehouses that still speak SQL, extensions for geo or vectors, teams that want one engine for transactional and moderate analytics workloads on a replica.
When to skip: Petabyte-scale warehouse analytics where a columnar EDW (BigQuery) is already standard. Document-only workloads with no relational joins (NoSQL patterns). Legacy MySQL apps you are not migrating yet.
Key features: MVCC, robust indexing, foreign data wrappers, row-level security, mature backup and HA patterns.
Details
| Topic | Notes |
|---|---|
| Deploy | Managed (RDS, Cloud SQL, Neon, Supabase) or self-hosted; Helm charts for K8s |
| Ops | pg_dump, PITR, connection pooling via PgBouncer or pooler in managed offerings |
| BI | Native connector in Grafana, Metabase, Redash, Apache Superset |
Practices: Migrations with Flyway or similar; read replicas for BI; separate roles for app vs analyst; prefer Postgres over MySQL for greenfield relational work.
References