Why we replaced analytics on Mongo with StarRocks
What broke when analytical queries shared a tenant cluster with OLTP — and what moving to a real OLAP engine actually gave us.
// data platform
Mongo → StarRocks
ms-latency dashboards
Stub. Full post body to follow.
Analytical queries on a multi-tenant Mongo cluster queue behind the transactional write load. We moved the analytical side onto StarRocks, fed from Kafka via Debezium CDC, and got ms-latency dashboards as a side effect.
// RELATED
2026-03-20 · 11 min · DISTRIBUTED SYSTEMS
Building a CDC pipeline with Debezium — without breaking the source
Lessons from tailing a multi-tenant Mongo oplog into Kafka without putting the source database under more load than it already had.
2026-02-18 · 10 min · DATA PLATFORMS
A lakehouse with Iceberg and Airflow — bronze, silver, gold
Why the three-tier lakehouse pattern works for replacing OLAP-on-OLTP, and what each tier actually does in practice.