What's new from Grafana Labs

The revamped query editor experience is now in public preview. Since the initial private preview release, we’ve continued to refine the experience and added two new capabilities to help you work with complex panels even faster: multi-select with bulk actions and a stacked view.

Analyze metrics cardinality history directly in the Cost Management and Billing App

We’re excited to announce the general availability of Historical Cardinality!

When active series counts increase, it can be difficult to determine what changed and which metrics or labels are responsible. Historical Cardinality helps you investigate cardinality growth over time, identify the source of cardinality spikes, and uncover opportunities to reduce observability costs

You can now add static labels to Azure metrics credentials in Cloud Provider Observability. Static labels are custom key-value pairs that are attached to every metric collected from a credential, giving you a consistent, queryable dimension that Azure Monitor doesn’t provide on its own.

Find the trace you need, even when it was dropped. Adaptive Traces now keeps every dropped trace queryable by trace ID for 24 hours after ingest, so a trace ID from a log line, exemplar, or an upstream system always returns a result.

Recently, we introduced several improvements to annotations to make them easier to explore, navigate, and manage at scale, including annotations clustering and indicator controls. These changes were introduced in public preview and now they are generally available.

In Grafana Cloud, we’re rolling out a change to what resource administrators can see on a resource’s permissions list. If you can view and manage the permissions on a folder or dashboard, you will soon see every user and team that has access to it.

Grafana Cloud users are regularly updated with the latest, most secure version of Grafana. However, plugin lifecycle management has been a responsibility for server administrators. Plugins becoming outdated risks incompatibility with the running Grafana version and causing instability. They may have security vulnerabilities, or simply lack the latest features to help users be successful.

Earlier this year we announced the addition of section-level variables for rows and tabs. This feature is now generally available.

If your dashboard includes both an API gateway and a database, changing something like an $instance variable affects all panels at once. To avoid this, you might have to split services across separate dashboards, which defeats the goal of having a unified view.

Grafana Alerting’s provisioning endpoints now reject duplicate secret keys that differ in casing. Requests that include duplicate keys will now be treated as validation errors with a 4XX status code.

Get started optimizing your Cloud Metrics quickly and more easily.!

The Adaptive Metrics UI has been revamped to improve readability and instill confidence in our optimization recommendations. Here are a few of the highlights:

When an error appears in your app, you want to know which code change caused it — without digging through git log. Suspect Commits in Frontend Observability now pins the exact commit your app was built from at the top of the candidate list on the error summary page, so you can go straight from a stack trace to the change that introduced it.

After you’ve enabled the Kubernetes manifests feature, the k8s-manifest-tail component collects changes to resources on your Cluster. With this, you can these techniques to relate deployment changes to behavior you observe in Kubernetes Monitoring.

The Kubernetes Monitoring app runs health checks and notifies you on any detail page’s overview tab. This check surfaces problems in plain language before you even go looking.

Grafana Fleet Management recently made it easier to build Alloy configuration pipelines with the component editor. Now we’re introducing an additional tool to help you understand and manage your pipelines.

Grafana Cloud Synthetic Monitoring now gives you control over which k6 major version runs your scripted and browser checks. With k6 version channels, you can stay on a stable release while k6 continues to evolve, and move to a new major version on your own terms.