Elastic Delivers Industry-Leading Metrics Capabilities with Native Prometheus Support and Agentic Investigation Experience

Native PromQL, out-of-the-box Kubernetes agentic investigation, and automated migration from Datadog and Grafana, all on the platform many SREs already use for logs.

Elasticsearch, the Search AI company (NYSE: ESTC), today announced a wave of new capabilities that bring the same scale, performance, and operational simplicity of Elastic’s trusted platform, already highly regarded for logs, to the realm of metrics data. With native support for Prometheus and PromQL, out-of-the-box Kubernetes investigation workflows, and automated migration from Datadog and Grafana to Elasticsearch, Elastic now delivers a unified platform for both metrics and logs. Built on Elasticsearch’s columnar metrics engine, the platform can query metrics up to 30x faster and store data up to 2.5x more efficiently than Prometheus, with no cardinality limits or penalties for custom metrics.

The metrics landscape has changed dramatically. Kubernetes and microservices have already expanded the time-series data targeted by observability systems from thousands to millions of data points. AI workloads are now accelerating this expansion, making metrics not just a scalability problem, but a strategic cost and reliability problem. For most platforms, this expansion comes with a significant cost. Premium vendors increase costs with higher cardinality, while lower-cost alternatives force metrics and logs into separate backends and query languages. This often leads to reduced data collection to control costs, leaving engineers with less context to understand what happened when an incident occurs.

Elastic Observability addresses both problems on a single platform, storing OpenTelemetry, Prometheus native, and application-defined metrics alongside logs and traces at full resolution. No separate backends are needed, and there are no trade-offs in data retention. This release targets the metrics engine that powers this platform and the various capabilities that run on top of it.

Native PromQL and Prometheus Remote Write: Run PromQL queries natively in Kibana, and ingest Prometheus metrics via Remote Write, so you don’t need to change your existing dashboards, alert rules, or scrape configurations.

Out-of-the-box Kubernetes workflows and content: With ready-to-use agentic workflows, alert templates, machine learning (ML) anomaly detection jobs, and pre-built dashboards that launch when data is ingested into Kubernetes, SREs can go directly from alert to root cause investigation. SRE teams can derive value without building infrastructure from scratch.

Agentic investigation: When an alert fires, Elastic uses workflows, including ML anomaly detection, to correlate metrics, logs, and traces that already share a single backend to reveal what changed and how severe the deviation is before the responsible party is even notified. The Observability MCP App and Agent Skills provide the same investigation capabilities to Claude, Cursor, VS Code, and all MCP-enabled tools.

Automated migration from Datadog and Grafana: The Observability Migration Platform automatically converts dashboards, alert rules, and PromQL queries to their Kibana equivalents, so you can migrate them without rebuilding.

“Elastic is already the trusted platform for logs at scale for many SREs, and now we are extending that superior scale, performance, and operational simplicity to metrics, giving customers up to 30x faster metrics queries than Prometheus, native Prometheus support, and a more predictable cost model for high-cardinality metrics,” said Baha Azarmi, General Manager of Observability at Elastic. “By handling all signals in a single backend, using a single query language, and enabling investigation before alerts are even sent, SREs have full context when they need it most, without the prohibitive costs that force them to limit the data they store.”

“As we migrate more applications to Kubernetes and expand our cloud footprint, our data volume is growing rapidly, and so is the internal need for more granular, high-cardinality metrics,” said Jeff Beagley, Manager of DevOps, SRE, and Cloud Engineering at Bass Pro Shops. “Elastic’s new metrics capabilities are helping us keep up with this data growth and get the insights we need. These capabilities, along with OpenTelemetry support, will help us achieve better performance and reduce costs while maintaining visibility into our increasingly complex architecture.”

“The improved metrics performance of Elasticsearch, native Prometheus support, and our LogsDB and incident response workflows have shortened our internal teams’ incident response times and established a more unified view across all signals without having to jump between multiple systems,” said Iosif Tournas, Leader of Cybersecurity & Elastic Platform at Eurowings Aviation. “These new metrics capabilities complement and enhance the millions of log events per minute and APM traces we are already processing with Elastic Observability. This unified view has not only reduced operational stress but also broken down silos between teams, shortening the time to detect and respond to various issues.”

Availability

The columnar metrics engine (TSDS), time-series data support in ES|QL, PromQL execution in Kibana, and ingestion via Prometheus Remote Write are generally available (GA). Out-of-the-box Kubernetes infrastructure content, including dashboards, alert templates, SLOs, and ML anomaly detection jobs, are also generally available. The Observability MCP App, Agent Skills, and Observability Migration Platform are in technical preview. All capabilities are available on Elastic Cloud, Elastic Serverless, and self-managed environments. While Datadog does not support on-premises environments and Grafana limits its most valuable features to hosted environments only, Elastic provides the flexibility to run observability workloads where you want, according to your data and operational requirements.

Additional Resources

Read the blogs:

Elasticsearch: From Best-in-Class Logs to Best-in-Class Metrics

Elasticsearch’s Columnar Store: Up to 160x Faster, Up to 6.6x More Storage Efficient

Updates to Elastic Observability Metrics Pricing

Elastic Observability Documentation

Elastic Observability Labs

This release is based on the English release announced on June 30, 2026 (local time) by Elastic (Headquarters: Mountain View, California, USA). Please refer to the original text (full version) here.

About Elastic

Elastic (NYSE: ESTC) is a Search AI company that integrates advanced search knowledge and artificial intelligence (AI) to transform data into answers, actions, and outcomes. Elastic’s Search AI platform is the foundation of its Search, Observability, and Security solutions, used by thousands of companies, including a majority of the Fortune 500. For more information, visit https://www.elastic.co/jp.

Elastic and associated marks are trademarks or registered trademarks of Elastic N.V. and its subsidiaries. All other company names and product names may be trademarks of their respective owners.

###