SQLx Patterns in 2026 — Rust + Postgres Without the ORM
Practical SQLx: query! macro vs runtime queries, migrations with sqlx-cli, transactions, FromRow, and the discipline that makes SQLx shine.
Practical SQLx: query! macro vs runtime queries, migrations with sqlx-cli, transactions, FromRow, and the discipline that makes SQLx shine.
Practical Postgres JSONB: GIN indexes, jsonb_path queries, partial indexes, when to denormalize vs normalize, and the perf gotchas.
Production Postgres partitioning: range / list / hash, when partitioning helps vs hurts, pg_partman automation, and pitfalls with foreign keys and indexes.
Choosing a Postgres replication topology: streaming for HA, logical for upgrades / cross-cluster, cascading for fanout, and pitfalls in each.
Production Postgres backups: pgBackRest setup, WAL archiving, PITR, restore testing, retention, and the patterns that work for self-hosted and managed.
Production hybrid search with Postgres alone: pgvector for semantic, tsvector for lexical, RRF fusion for combining, optional reranker. Performance, tuning, and patterns.
Why connection pooling is non-negotiable for Postgres apps. PgBouncer transaction-mode quirks, modern alternatives (pgcat, Supavisor, Hyperdrive), and the patterns from production.
Time-series patterns: vanilla Postgres + time partitioning, TimescaleDB hypertables and continuous aggregates, ClickHouse for OLAP scale, and the decision matrix.
How to use Postgres RLS for multi-tenant safety. Policies, session variables, performance considerations, and the patterns that make RLS a cheap defense-in-depth layer.
Why VACUUM matters, how MVCC creates bloat, autovacuum tuning, bloat detection, pg_repack, and the playbook for keeping Postgres healthy at scale.