PostgreSQL

Posts on PostgreSQL — the database I reach for by default. Practical writing on the parts of Postgres that backend developers use every day: data modeling, indexing, query plans, JSONB, transactions, full-text search, and how to get from a slow query to a fast one.

Postgres Partitioning in 2026 — Range, List, and the Patterns That Scale

When and how to partition Postgres tables. Range / list / hash, partition pruning, attach/detach, default partitions, and the gotchas at scale.

Postgres Replication and HA in 2026 — Streaming, Logical, and Auto-Failover

Production HA for Postgres in 2026. Streaming vs logical replication, Patroni vs pg_auto_failover vs Stolon, RPO/RTO tradeoffs, and the patterns that keep a primary alive.

Postgres CDC and Logical Replication in 2026 — Debezium, pgoutput, and the Outbox Alternatives

How CDC actually works in 2026. Postgres logical replication slots, pgoutput, Debezium, the outbox pattern, and how to fan out changes to Kafka, NATS, ClickHouse, OpenSearch, or your AI pipeline.

Distributed SQL in 2026 — CockroachDB, Spanner, Yugabyte, and When You Actually Need One

A practical 2026 guide to distributed SQL. The architecture behind CockroachDB / Spanner / Yugabyte / TiDB, when distributed SQL earns its cost, when Postgres + read replicas is enough, and the migration realities.

PostgreSQL MVCC, Isolation, and Locking — A Backend Developer's Guide

How Postgres MVCC actually works, the four isolation levels (and what each gives you), row vs advisory locks, deadlocks, and the transaction patterns every backend developer needs to handle concurrency correctly.

PostgreSQL 18 — Async I/O, UUIDv7, Virtual Generated Columns, and What Actually Matters

A focused walkthrough of PostgreSQL 18 for application developers — async I/O, native uuidv7(), virtual generated columns, OAuth auth, EXPLAIN buffer reporting by default, logical replication improvements, and what to know before you upgrade.

PostgreSQL JSONB Advanced Patterns — Indexing, Performance, and Pitfalls

JSONB beyond the basics — GIN with jsonb_path_ops, partial and expression indexes, generated columns for hot fields, partitioning, jsonb_set rewrites, and the mistakes that turn 12ms queries into 4 seconds.

pgvector Deep Dive — HNSW, IVFFlat, and Tuning Postgres for Vector Search

Everything you need to make pgvector fast in production: HNSW vs IVFFlat, distance operators, m and ef_construction, ef_search at query time, partial indexes for multi-tenant data, and benchmarking your real workload.

PostgreSQL Full-Text Search Without an Extra Service

Postgres ships with surprisingly capable full-text search built-in. Here’s how to use it for ranked text search without adding Elasticsearch to your stack.

PostgreSQL Indexing and EXPLAIN: Make Slow Queries Fast

Learn to read EXPLAIN ANALYZE output, pick the right index type for the query, and turn slow Postgres queries into fast ones — without guessing.