Onboarding onto a legacy schema

Situation

Your team just inherited a database. No docs, no idea which of the 40 tables are load-bearing, and a suspicion that several decade-old column types are costing you.

Commands

pgdba check --connection legacy
pgdba check --only schema.design.keys,schema.design.types,schema.design.normalization
pgdba check --explain-term "rewrite lock"

What pgdba does

The schema-design family audits keys (missing/duplicate PKs, FK coverage), sequence hygiene (identity vs serial, overflow headroom), collation consistency, column types (wasteful CHAR/TEXT/NUMERIC patterns), table width, normalization smells, matview refresh strategy, unlogged tables, and index depth — each finding ranked with evidence.

Done looks like

A prioritized modernization to-do list in the first week (the evidence page shows what fixing the column-type class is worth: −54% disk, −31% agg in the benchmark), instead of tribal knowledge assembled over a quarter.