Instrumentation Hub in Grafana Cloud: start easier, stay in control

It's no secret that getting started with observability can be much harder than it should be. That's why, at ObservabilityCon 2025, we introduced Instrumentation Hub: a guided, scalable way to roll out observability coverage while still keeping you in control. 

Today, we’re excited to announce that Instrumentation Hub is now Generally Available in Grafana Cloud.

As we hinted at before, this new iteration introduces selective OpenTelemetry SDK injection, giving you deeper runtime insights for a wide range of languages to complement eBPF instrumentation with a single click. 

What is the Instrumentation Hub?

Instrumentation Hub makes the biggest bottleneck of starting with observability, instrumentation, faster and easier by leveraging open source eBPF and OTel injection.

At Grafana Labs, we're champions of open-standards-based observability, but we also realize that getting meaningful coverage of your systems can often mean months of manual toil. Each layer requires its own instrumentation approach, balancing technology tradeoffs, and unique brittle steps. 

Organizational challenges exacerbate this process, making observability a moving target with uneven coverage, persistent blind spots, and ultimately leaving you wondering if the effort was worth the pain.

We believe there’s a better way. Instrumentation Hub is a control plane for remote discovery and selective auto-instrumentation. It helps you go from zero to meaningful telemetry in a matter of minutes by removing the bottleneck of having to apply instrumentation by hand.

How Instrumentation Hub works

The premise is simple: you run a single Helm command on your Kubernetes cluster to install Alloy, our distribution of the OpenTelemetry Collector.

Once installed, Instrumentation Hub automatically catalogs everything that’s running on the cluster, structured by namespace and technology. Teams can immediately see what exists, what’s instrumented, and what isn’t.

Then, you use either the Grafana Cloud UI, an API automation, or your favorite LLM agent to describe what and how you want to collect and monitor.

Under the hood, Instrumentation Hub translates these selections into observability pipelines without the need for manual configuration. Within minutes you can begin exploring insights across your infrastructure and services in the Grafana Cloud observability platform.

And of course, this isn’t a one-time setup. Teams can come back any time and refine their observability scope by namespace and enable or disable specific signal collection to balance their observability budgets. Need profiles for a misbehaving service? Enable them temporarily. Too many logs from a noisy workload? Exclude them. Trying to reduce spend? Dial back collection where it’s no longer needed.

Observability is a progression

Last year at ObservabilityCon, we introduced the idea of a hierarchy of observability needs as a mental model to chart an effective path for applying observability. We described how you should start from a bird’s eye view of your estate and gradually move towards deeper and more customized insights.

Instrumentation Hub is built around this progression, facilitating observability in the right order for a maximum return on investment.

At the base is the foundation everything depends on: infrastructure visibility, which provides insight into the machines the code runs on. Instrumentation Hub helps you get started at this layer by enabling Kubernetes Monitoring with the click of a button.

Above infrastructure is baseline service visibility. RED metrics and service relationships provide a consistent, system-wide view so teams can quickly triage issues and narrow the search space before digging deeper. Instrumentation Hub populates Grafana Cloud's Knowledge Graph in a matter of minutes, providing rich context for humans and agents.

At the next level, technology-specific and transaction-level insights become necessary. Distributed traces reveal application behavior across the full request path, exposing performance bottlenecks and dependencies that aren’t visible from the outside.

This is where Instrumentation Hub stands out. It can selectively deploy OpenTelemetry auto-instrumentation SDKs with the click of a button.

Historically, this process meant installing a new Operator, deploying instrumentation CRDs, annotating your target applications, and restarting the pods. This work can span across multiple teams, ownership boundaries, or require a dedicated new collection layer. But now, the very same upstream SDKs can be rolled out to complement and enrich your existing baseline service visibility without any extra moving pieces and without disruptive restarts.

Get started with Instrumentation Hub today

Try it out! The Instrumentation Hub is available to all Grafana Cloud stacks for your Kubernetes clusters running on Linux.

If you’ve recently installed the Kubernetes Monitoring Helm chart you might be set up with it already. And if you're still unsure, ask Grafana Assistant for help!

Miss out on ObservabilityCON 2026? Check out our blog post recapping all the big announcements!