ClickHouse performance tuning, backed by real query data
Find the queries actually costing you time and memory, then get recommend-only tuning suggestions from EXPLAIN and a capacity/TTL advisor — chmonitor never rewrites schema or executes DDL for you.
Built for this, not bolted on
Slowest and most expensive queries
Ranked by duration, memory and rows read, so you know exactly where tuning effort will actually pay off.
EXPLAIN-based suggestions, recommend-only
Missing PREWHERE, unpruned partitions, unkeyed joins and missing index hints, surfaced as suggestions you review and apply yourself.
Capacity & TTL advisor
Ask the built-in AI agent to forecast disk-full dates from write growth and suggest a TTL change to stay under budget. It returns a suggested ALTER TABLE ... MODIFY TTL statement — it never runs one.
Community
Open source, built in public
Related
More ways to monitor ClickHouse
Start monitoring — advisor and alerts built in
Open the hosted dashboard and connect a cluster, or self-host with Docker, Kubernetes, or from source. Self-hosting is always free.