8 major updates to Cloudflare Observability

Cloudflare is launching eight major updates to create a unified observability platform by consolidating logs, traces, analytics, alerts, dashboards, and exporting into a single system with simpler pricing. Key features include a centralized place to explore all logs across Cloudflare products, end-to-end tracing from the edge to the origin, a unified SQL API for querying data, and a new unified pricing model for ingested and stored observability data. Users can now configure custom alerts on this data, view domain analytics with 30 days of retention, build custom dashboards, and export data via Logpush, which is now available on self-serve plans.
The goal of these updates is to address the fragmentation of data across different Cloudflare products, aiming for observability that reflects real application behavior by providing complete context needed for issue resolution. The updates focus on making investigation more cohesive through unified querying methods like SQL and tracing capabilities. Furthermore, the platform is moving towards a future where more Cloudflare products integrate into this shared observability capability with consistent pricing and features.

* Eight major updates are being launched for Cloudflare Observability.
* A single place exists to explore logs across Cloudflare, combining Workers Observability and Log Explorer.
* End-to-end tracing from the Cloudflare edge to the origin is being introduced in open beta.
* A unified SQL API is launching in beta for querying telemetry across Cloudflare data.
* A unified pricing model is established for observability data ingested and stored across Cloudflare, effective December 1, 2026.
* Custom alerts can be configured on observability data via the new SQL API.
* Domain analytics are now available with 30 days of retention across all plans.
* Logpush export functionality is now available on self-serve plans.
* Logs include datasets for HTTP events, firewall events, Workers, Containers, R2, and AI Gateway.
* Observability data can be queried using raw SQL or built-in filters.

The narrative is shifting from siloed product monitoring to a platform-wide observability capability, driven by the need for context across disparate systems that govern application behavior. The introduction of unified querying via SQL and end-to-end tracing positions Cloudflare not just as an infrastructure provider, but as the contextual layer necessary for operational reality. This move fundamentally addresses the challenge faced by operations teams: correlating signals from various components (Workers, origin, network) to diagnose complex issues.
The pattern observed is the structural alignment of data management: consolidating metrics, logs, and traces under a unified framework, backed by standardized interfaces like OpenTelemetry and SQL. The implication is that true operational efficiency relies on standardized context propagation and unified access layers, suggesting a maturation in how infrastructure observability is conceptualized. The shift to usage-based pricing for ingestion, combined with extended retention options, suggests an acknowledgment of long-term trend analysis as a core requirement rather than just immediate incident response.
What assumptions underpin the necessity of this consolidation? It assumes that operational teams require this level of cross-product context consistently. The architecture favors abstraction—using concepts like tracing and SQL to link disparate data sources—suggesting the underlying goal is to empower agents by providing portable interfaces (OpenTelemetry, W3C Trace Context) to operate effectively in a complex environment. The critical question becomes whether this unified layer truly facilitates deeper systemic understanding or merely creates another interface layer over existing complexity. How does the focus on agent empowerment reconcile with the proprietary nature of the resulting platform?