PostgreSQL

Star history isn't available yet — tracking begins with the next ingestion run.

A general-purpose, client-server relational database with a three-decade track record in production.

What it does

PostgreSQL is an object-relational database management system that traces its lineage to the POSTGRES project at the University of California, Berkeley, started in 1986; the modern open-source project (adding SQL support and taking the name "Postgres95", later PostgreSQL) began in 1996 after the original Berkeley team moved on. It runs as a long-lived server process that clients connect to over a network or local socket, rather than linking into an application the way an embedded database does, and it's built around strict SQL standards compliance plus deep extensibility — custom data types, functions, and index types (used by extensions like PostGIS for geospatial data) can be added without forking the database itself.

Why people use it

• Full ACID transactions with mature, well-understood concurrency control (MVCC), so many clients can read and write at once without blocking each other
• Rich, standards-compliant SQL plus advanced data types (JSON/JSONB, arrays, ranges, full-text search) beyond what a minimal SQL engine offers
• A large extension ecosystem (PostGIS, pgvector, TimescaleDB, and others) that adds specialized capabilities directly inside the database
• The default or supported choice for most managed cloud database services, which keeps operational tooling, drivers, and hosting options broad

Pros

• Decades of production hardening across every major workload shape, from small apps to very large multi-tenant systems
• Genuinely extensible at the SQL and type-system level, not just through plugins bolted on from outside
• Strong data-integrity guarantees (constraints, foreign keys, strict typing) enforced by the database itself, not left to the application

Cons

• Needs a server process to install, configure, tune, and keep running — real operational overhead compared to an embedded database with no service to manage
• Vertical scaling (one larger machine) is the primary scaling path; horizontal write-scaling across nodes isn't built in the way some distributed databases offer
• Heavier to spin up for a quick script, a mobile app, or a single-file local datastore — see SQLite below when a server isn't worth running

Related Grove

Databases