Grafana Mimir 3.0 Launch Expands Open Observability at Scale

ByRicardo

New Tempo launch brings AI-assisted tracing; Kubernetes Monitoring provides fleet administration and Helm Chart v2; Mimir 3.0 preview expands open observability at scale.

Grafana Labs, the corporate behind the open observability cloud, at present introduced the launch of Grafana Mimir 3.0, the newest evolution of its open-source, horizontally scalable metrics backend. Introduced at KubeCon and CloudNativeCon North America 2025, Mimir 3.0 marks an architectural milestone, delivering new ranges of reliability, efficiency, and value effectivity for Prometheus-compatible monitoring at enterprise scale.

Built on open supply, open requirements, and open ecosystems, Grafana Labs helps organizations innovate with out lock-in and transfer quick with out compromise. At KubeCon, the corporate additionally introduced updates throughout its open-source ecosystem, together with Grafana Tempo 2.9 with AI-assisted tracing, continued Kubernetes Monitoring enhancements, and deeper Prometheus and OpenTelemetry help to assist groups simplify observability and achieve extra worth from their knowledge.

“From open supply and open requirements to open ecosystems and open minds – constructing within the open is core to our philosophy at Grafana Labs,” stated Myrle Krantz, Senior Director of Engineering, Grafana Labs. “That’s why we’re persevering with to spend money on open supply, like including AI-assisted tracing in Tempo and making it simpler to get essentially the most out of OpenTelemetry and Prometheus. We’re constantly bettering Kubernetes Monitoring. And with the brand new Mimir 3.0 launch, we’re serving to groups scale much more reliably, increasing what’s potential for open observability in 2026 and past.”

Grafana Mimir 3.0: Improved Reliability and Performance

Three years in growth, Grafana Mimir 3.0 introduces a brand new decoupled structure that separates the learn and write paths for extra dependable, large-scale metrics operations:

  • Reliability: By decoupling reads and writes by way of an asynchronous Kafka-based ingest layer, cross-path dependencies are eradicated, holding queries quick and secure even beneath heavy ingestion masses.
  • Performance: The new Mimir Query Engine (MQE) streams question outcomes as an alternative of loading whole datasets into reminiscence, bettering execution velocity and decreasing reminiscence utilization by as much as 92%.
  • Cost effectivity: Early testing stories as much as 15% decrease useful resource utilization whereas attaining greater throughput and consistency throughout giant clusters.

Together, these improvements make Mimir 3.0 essentially the most resilient, high-performing, and cost-efficient metrics backend for Prometheus and OpenTelemetry knowledge – now accessible on Grafana Cloud and for self-managed customers through open supply.

Grafana Tempo 2.9: AI-Assisted Tracing and Faster Queries

The newest launch of Grafana Tempo, the open supply distributed tracing backend, introduces new capabilities to hurry up hint evaluation and convey AI into the observability workflow.

  • MCP server help: An experimental Model Context Protocol (MCP) server permits AI assistants like Claude Code and Cursor to question distributed tracing knowledge with TraceQL, enabling natural-language debugging and quicker root trigger evaluation.
  • TraceQL metrics sampling: New probabilistic question hints speed up evaluation in high-volume environments, returning approximate outcomes quicker with out dropping visibility.
  • Multi-tenant and operational enhancements: New metrics for question I/O, span timing, and utilization monitoring enhance observability and efficiency visibility at scale.

Tempo 2.9 additionally deepens OpenTelemetry help by aligning with newer OpenTelemetry semantic conventions, reaffirming Grafana Labs’ dedication to open, composable observability.

Kubernetes Observability: Intelligent, Automated, and Built for Modern Workloads

Building on the success of Grafana Cloud Kubernetes Monitoring, Grafana Labs has launched highly effective new capabilities that simplify observability throughout even essentially the most advanced Kubernetes environments. This is very well timed as a latest survey by the CNCF discovered that 80% of respondents work for IT organizations which have deployed Kubernetes in a manufacturing surroundings.

Kubernetes Monitoring in Grafana Cloud has developed right into a best-in-class, opinionated observability answer that doesn’t simply visualize telemetry however interprets it, automates insights, and guides groups to motion. New updates embrace:

  • Grafana AI Assistant integration: Teams can now work together with Kubernetes Monitoring utilizing Grafana Assistant (now typically accessible), an AI-powered agent constructed into Grafana Cloud that may learn dashboards, drill into panels, and summarize leads to actual time. Using pure language, customers can ask how a workload is behaving, what’s impacting efficiency, or the place prices are trending.
  • GPU monitoring: Available at each the Node and Cluster stage, new GPU utilization panels assist detect overheating, energy drain, or underuse earlier than they influence efficiency, making certain AI workloads stay secure and environment friendly.
  • Automated root trigger evaluation: Now built-in with the commonly accessible Grafana Knowledge Graph, Kubernetes Monitoring provides you computerized RCA and Insight Rings.
  • Expanded workload help: Kubernetes Monitoring now offers full visibility into CronJobs, Argo Rollouts, Bare Pods, Static Pods, Strimzi Pod Sets, and different nonstandard workloads, making certain complete protection throughout various infrastructure varieties.
  • Monitor cron jobs and different job varieties: Get full visibility into all cron and handbook jobs throughout clusters. Instantly see standing, distribution, and missed runs to make sure automation reliability and fast difficulty detection.
  • CPU and reminiscence panels: New CPU and Memory tabs present clear, layered views of compute utilization – from cluster to container – with effectivity graphs and CPU distribution evaluation that assist optimize capability, price, and efficiency.
  • Cloud supplier nodes: One-click correlation between AWS EC2 situations and Kubernetes workloads permits unified troubleshooting throughout cloud and container layers, decreasing context-switching and imply time to decision. And for groups on AWS, CloudWatch metric streams in Grafana Cloud can reduce metric pipeline prices, together with storage and agent infrastructure, by as much as 10x whereas delivering near-real-time metrics.

Together, these updates make Kubernetes Monitoring in Grafana Cloud an clever, automated, and AI-capable answer for at present’s dynamic, large-scale environments.

At the latest ObservabilityCON occasion in London, Backbase methods engineer Andrei Drumov stated of Kubernetes Monitoring, “We are amassing the telemetry alerts utilizing Grafana Alloy and we’re utilizing the Kubernetes Monitoring Helm charts developed and maintained by Grafana to handle the Grafana Alloy deployments. We selected this path due to its simplicity, flexibility, and modularity. It’s comparatively straightforward to handle. It permits us the flexibleness to increase the configurations in case we want customized integrations, and it additionally permits us to conditionally allow or disable sure options based mostly on sure buyer wants, as a result of some clients may have software availability, another clients may have steady profiling, some clients may have each. And with Kubernetes Monitoring Helm charts, it’s very easy to juggle.”

Open Source Ecosystem Highlights

According to Grafana Labs’ 2025 Observability Survey, open requirements like Prometheus and OpenTelemetry proceed to achieve momentum throughout the business. The survey discovered that 71% of organizations use each OpenTelemetry and Prometheus, with greater than half growing their investments in OpenTelemetry for the second yr in a row. While 67% of organizations use Prometheus in manufacturing in some capability, OpenTelemetry is on a powerful development trajectory, with 38% of respondents investigating adoption and solely 6% reporting no plans to make use of it at all. The findings additionally present that vendor neutrality and suppleness stay essentially the most cited necessities for observability instruments, instantly aligning with Grafana Labs’ open, composable method.

Building on these tendencies, Grafana Labs continues to spend money on open requirements and community-led innovation throughout its ecosystem:

  • Beyla donation full: Earlier in 2025, Grafana Labs donated Grafana Beyla, its eBPF-based, zero-code auto-instrumentation agent, to OpenTelemetry. Renamed OpenTelemetry eBPF Instrumentation, the mission simply marked its first official launch beneath the OpenTelemetry umbrella. The donation reinforces Grafana Labs’ long-standing dedication to advancing open, vendor-neutral observability.
  • Grafana Alloy: Grafana Labs’ distribution of the OpenTelemetry Collector, Grafana Alloy is now the default knowledge pipeline layer throughout Grafana Cloud and open supply deployments. Alloy unifies metrics, logs, and traces assortment whereas supporting each Prometheus and OpenTelemetry pipelines.
  • Prometheus 3.0 and OpenTelemetry interoperability: Grafana engineers contributed to the introduction of profiling sign help, new semantic conventions, and Prometheus 3.0 compatibility, strengthening cross-project interoperability.

See Grafana Labs at KubeCon + CloudNativeCon North America 2025

Visit Grafana Labs sales space #444 for stay demos and listen to extra about Grafana Labs’ merchandise and open supply contributions throughout the next periods that includes Grafana Labs engineers:

  • Monday, November 10 at 10:40 am EST: Telemetry That Matches Your Model, Verified Live – Liudmila Molkova, Staff Developer Advocate, Grafana Labs
  • Tuesday, November 11 at 5:45 pm EST: Straight Into the Deep End! Learning Kubernetes and Cloud Native From Scratch in Late 2024 – Éamon Ryan, Senior Principal Field Engineer, Grafana Labs and Ayah Elshaikh, Senior Field Engineer, Grafana Labs
  • Wednesday, November 12 at 2:15 pm EST: UX Research Report: Prometheus and OTel’s Resource Attributes – Amy Super, Principal Product Designer, Grafana Labs
  • Wednesday, November 12 at 4:00 pm EST: Debugging your cluster when it’s on fireplace – Nikola Grcevski, Principal Software Engineer, Grafana Labs
  • Wednesday, November 12 at 5:30 pm EST: Observing Dark Matter with OpenTelemetry – Mario Macías, Staff Software Engineer, Grafana Labs
  • Thursday, November 13 at 1:45 pm EST: Prometheus Intro, Deep Dive, and Open Q+A – Owen Williams, Principal Software Engineer, Grafana Labs

The put up Grafana Mimir 3.0 Launch Expands Open Observability at Scale first appeared on AI-Tech Park.

Similar Posts