ClickHouse Embraces PromQL: A Prometheus Power-Up

PromQL in ClickHouse: A Seamless Migration Path

ClickHouse's introduction of PromQL support and a dedicated TimeSeries engine marks a pivotal moment for observability data management. The core innovation lies in its 'drop-in replacement' approach, allowing existing Prometheus users to leverage their current PromQL queries and tooling directly within ClickHouse Cloud. This significantly lowers the barrier to entry for migrating metrics infrastructure, eliminating the painful process of rewriting queries in SQL and the associated risk of semantic drift. By handling the complexities of Prometheus's unique data model and query language semantics server-side, ClickHouse offers a compelling proposition for organizations looking to consolidate their observability data (logs, traces, and now metrics) into a single, performant platform.

The implementation details, such as the TimeSeries table engine built on ClickHouse's storage engine, are crucial. The blog post hints at deep technical work to ensure PromQL semantics are preserved, which is a non-trivial task given Prometheus's history and design decisions around data presentation, scraping, and encoding. The promise of 'transpil[ing] PromQL into performant SQL' is a strong technical claim that will be closely watched. The private preview offering remote write ingestion and query via ClickStack, Grafana, and clickhouse-client provides immediate utility. However, the exclusion of AlertingRules and RecordingRules from the current preview, while slated for future announcements, means a complete Prometheus replacement is not yet realized, which could be a limitation for some early adopters.

This move directly challenges existing Prometheus solutions and potentially other time-series databases. The ability to query metrics using familiar PromQL syntax alongside logs and traces within ClickHouse provides a unified operational data store. The focus on performance and the promise of 100% PromQL coverage are ambitious and will be key differentiators. The strategy of collecting user feedback on unsupported queries is a smart way to prioritize future development. The potential for unifying OpenTelemetry metrics with PromQL in the future further solidifies ClickHouse's ambition in the observability space.

Key Points

• ClickHouse introduces a new TimeSeries table engine and PromQL support in ClickHouse Cloud, aiming to be a 'drop-in replacement' for Prometheus.
• Users can send Prometheus metrics via remote write and query them using existing PromQL queries in ClickStack, Grafana, or ClickHouse.
• The integration preserves PromQL semantics, eliminating the need for users to rewrite queries in SQL and ensuring consistency.
• The private preview covers metrics storage and querying, while collection remains handled by existing Prometheus servers or collectors.
• Future plans include support for AlertingRules, RecordingRules, Native Histograms, and OpenMetrics/remote-write v2 protocols.

---

📖 Source: Introducing ClickHouse's new TimeSeries Engine: Your drop-In Prometheus replacement