Every query, explained
Running, history, failed, slowest and most expensive queries — duration, memory, rows read and the full statement, in sortable tables backed by system.query_log.
What is running right now
A live table of every executing query with elapsed time, memory, rows read and the user behind it. Long-runner misbehaving? Kill it from the row menu without opening a client.
- Queries by user and top patterns over time
- Saved filter presets for repeat triage
- Full statement with SQL formatting on demand
A year of activity at a glance
The query heatmap shows volume across a full year, and record-breaker insights surface the slowest, hungriest and most expensive statements your cluster has ever seen.
- Heatmap of query volume by day
- Record-breakers: duration, memory and rows-read champions
- Failed queries with error classification
From symptom to plan in one click
Open EXPLAIN for any statement to see the plan, the indexes used and the parts scanned — then hand it to the AI agent for a rewrite recommendation.
- EXPLAIN plan and pipeline views
- Memory breakdown per query
- One-click handoff to the AI agent
More of Query monitoring in action
Everything in the box
Running queries
Live table with elapsed, memory, rows and kill action.
Slow & expensive
Slowest and costliest statements ranked from query_log.
Failed queries
Errors grouped and classified for fast triage.
EXPLAIN
Plan and pipeline views for any statement.
Heatmap
A year of query volume in one visual.
By user
Who runs what — usage split by user and pattern.
Common questions
Does query monitoring need an agent on my hosts?
No. Everything reads ClickHouse system tables (query_log, processes) over the native HTTP interface — no sidecar, no agent.
Can I kill a query from the dashboard?
Yes — the row action issues KILL QUERY. Control actions are permission-gated and can be disabled entirely.
Which ClickHouse versions are supported?
Queries are version-aware: chmonitor picks the right SQL for your server version, back to 23.x system table schemas.
More chmonitor features
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.