Loading post…
ClickhouseAnnouncement
Announcing the Managed ClickStack MCP Server
Summary
Managed ClickStack in ClickHouse Cloud now includes the ClickStack MCP server, extending its observability investigation capabilities to external agents and AI workflows. Unlike the generic ClickHouse MCP server’s SQL-oriented interface, it offers semantic tools for log patterns, outlier correlation, slow traces, and repeatable investigations across logs, metrics, and traces. For event patterns, the server samples matching events, processes them with Drain3 in the MCP layer, and extrapolates and ranks counts, reducing output cardinality and making query overhead fixed rather than linear in the full dataset. Internal benchmarks report 25% fewer tool calls, 2.5x greater consistency, and an almost 20% increase in evaluation scores versus the standard ClickHouse MCP. Managed deployments also support full-fidelity telemetry, dedicated compute for agent workloads, and orchestration such as persisted dashboards, while native SQL remains available for unusual or deeper questions.
Context
Observability investigations often require multi-step, domain-specific analysis across logs, metrics, and traces. The post says that repeatedly generating raw SQL makes models spend context reconstructing query patterns and analysis logic instead of investigating the underlying problem, while large event-pattern queries can create high-cardinality outputs, long runtimes, or unnecessary server pressure.
Approach / What changed
The Managed ClickStack MCP server exposes higher-level semantic investigation tools that execute optimized ClickHouse queries underneath. Its event-pattern workflow uses a small number of lighter queries, including a random sample and a total-result count, then processes the sample with Drain3 and extrapolates and ranks the final counts. Managed deployments can also use dedicated compute for agent workloads and persist investigation results through dashboards and other orchestration primitives.
Takeaways
The specialized tools cover recurring log-pattern mining, time-window comparisons, trace-outlier root cause analysis, and investigations that move across logs, metrics, and traces without requiring the model to reconstruct each workflow from raw SQL.
A generated MCP connection string contains a ClickHouse Cloud service ID that selects the executing service. Launching ClickStack from a child service allows agent workloads to run on dedicated compute, isolated from ingestion and user-facing workloads.
Operational workflows can create dashboards, persist searches, manage alerts, and share findings, turning a local agent investigation into evidence that can be reviewed by a wider team rather than leaving it in raw chat output.
Source and freshness
- Source
- clickhouse.com
- Published
- Added to Yomu
- Last updated
Summaries are generated from the original by a language model. Check the original before relying on it.
Keep reading
Related reading
The Good Stuff
Paid picks we use ourselves. Links carry a referral code — a subscription through one earns us a commission at no extra cost to you.
The best engineering reads, once a week
One useful email. Unsubscribe anytime.