Data infrastructure, measured on my own hardware.

No vendor decks, no reprinted docs. I rebuild the same pipeline in every tool I write about, break it on purpose, and publish the harness next to the numbers.

Method 8 min read

The DeWitt clause

Why database licences spent forty years forbidding you to publish a benchmark, and what that did to the evidence.

Ingestion 10 min read

The diskless-Kafka claim, priced

Cross-AZ replication set against S3 request costs, from published rate cards. No rig, no latency figures — arithmetic only.

Method 8 min read

What an audited benchmark looks like

LDBC has independent certified auditors read the code and run the thing. Almost nothing else in this field does.

Method 8 min read

What ClickBench doesn't measure

Its 43 queries touch one table and never join. Here is the method this publication runs under, argued from that.

Method 9 min read

Who pays for the numbers you cite

A survey of where data-infrastructure benchmarks come from, and how often the answer is a company selling the winner.

Orchestration 3 min read

Airflow, Dagster, Prefect: figures withdrawn

This page carried scheduler numbers that came from the site's own build fixtures. They were never measured, and they are gone.

Ingestion 3 min read

CDC without the Kafka bill: figures withdrawn

This page carried a change-data-capture comparison, and two corrections to it. All three were build fixtures, never measurements.