Keitaro on ClickHouse vs MySQL — pick your storage engine by volume

Keitaro on ClickHouse vs MySQL — pick your storage engine by volume

Keitaro can run clicks on MySQL or ClickHouse. This is an infrastructure decision that decides how high you can scale before reports crawl.

Use MySQL when:
— Under ~1-2M clicks/day. Simpler ops, one less service.
— You run frequent small UPDATE-style operations.
— Your VPS is RAM-constrained — ClickHouse is hungry.

Use ClickHouse when:
— High volume — it eats hundreds of millions of rows for analytical queries.
— Your report builder times out on MySQL at current volume.
— You want fast group-by across long date ranges.

Tradeoffs to log:
— MySQL handles writes + reports adequately at low volume, chokes on big aggregations.
— ClickHouse flies on reads/aggregations but needs more RAM and careful setup.
— Migrating mid-scale means moving live click data — plan a maintenance window.

Rule: start MySQL, watch report-build time. When aggregate reports cross ~5s consistently, plan the ClickHouse move before you're forced to.

Save this SOP. Run this before every launch.