Blog

Practical writing on running PostgreSQL well - performance, reliability, and knowing what to fix first.

PostgreSQL logging

The logging setting that slows down new connections

October 2026 · 12 min read

The client address in PostgreSQL's log comes from %h, not from log_hostname. What log_hostname adds is a DNS lookup at the start of every new connection. We explain how the two settings work together and what the lookup cost when DNS was healthy, slow and down.

Read the article
PostgreSQL tuning

max_wal_size, explained

July 2026 · 9 min read

max_wal_size gives PostgreSQL's transaction log a size budget. Too small for the workload, and the server writes the same data to disk again and again. We explain how the mechanism works, what our internal benchmark measured, how to tune it with your server's own counters, and when to leave it alone.

Read the article
Manifesto

Why we're building Guardioz

June 2026 · 5 min read

PostgreSQL keeps the business running. When it slows down, someone has to find out why - and it's rarely a dedicated DBA. A short manifesto on the gap between a symptom and a decision, and why we're building the guardian that closes it.

Read the article