Recent changes and improvements to Dash0

We've increased the retention period for synthetic API check runs from 30 days to 13 months at no extra cost. This change applies to all check runs executed from January 19, 2026 onwards. Existing check runs will continue to follow the previous 30-day retention policy.

Starting with version 0.99 of the Dash0 Kubernetes Operator, you can define per-namespace settings for exporting telemetry and syncing custom resources (Dash0ViewCopy, Dash0SyntheticCheckCopy, PrometheusRuleCopy, PersesDashboardCopy) to different backends and/or datasets.

The configuration in the Dash0OperatorConfigurationCopy serves as a default that can be overridden by providing an alternative configuration in a namespace's Dash0MonitoringCopy resource.

You can find more details in the operator's documentation.

We've added five new AI integrations to help you build and monitor intelligent applications:

AI Coding Assistants (connect to your Dash0 telemetry):

• OpenCode - Terminal AI coding agent that uses your telemetry data to identify issues and correlate incidents with code changes.
• Kiro - Agentic IDE with spec-driven development that leverages your observability data for telemetry-aware coding and root-cause analysis.

OpenTelemetry Monitoring (observe your AI applications):

• kagent - Kubernetes-native platform for building and deploying AI agents that can connect to MCP servers (including Dash0). Monitor your agents' LLM calls, tool executions, and orchestration flows with built-in OpenTelemetry support.

LLM Application Monitoring (observe your AI applications):

• LangChain - Monitor LLM call performance, token usage, and chain execution flows in your LangChain applications.
• OpenLIT - OpenTelemetry-native observability for LLM applications with automatic instrumentation and GPU monitoring.

These integrations enable you to build AI-powered tools that understand your system's behavior through telemetry, and monitor the performance and costs of your AI applications in production.

Dash0 Folders let you organize assets by service, environment, or business domain, reflecting how your team actually works. This makes it easier to find the right dashboard and keeps your workspace organized as your system evolves.

Folders are built for scale. You can manage them through Infrastructure as Code, keeping your organizational structure consistent across development, staging, and production. Teams can also define which folders appear by default, giving everyone a clean and intentional starting point.

In short: Folders reduce clutter and bring structure to growing observability environments, so Dash0 remains as organized as the systems you monitor.

📊 Service Catalogue

Get a bird’s-eye view of your system health or dive deep into specific code paths.

• The Command Center: A high-level list of every service.
• Infrastructure Resources: See "Failed Checks" on Resources alongside your APM metrics.

Endpoint-Level Granularity: Use the Operations Tab to see which specific operation is failing, even when the rest of the service looks healthy.

🗺️ Service Map

We had a Service Map before so what's new?

• Customizable Views: Save your preferred filters to quickly see the services most relevant to your work by default.
• Dynamic Traffic Visualization: Easily discover traffic flows and paths within your service topology with overlaid particles.
• Grouped External Services: Explore external services organized by similarity, reducing clutter and complexity compared to the full internal map.

We use OpenTelemetry semantic conventions to automatically draw the connection between your services.

• Automatic Discovery: Connections are derived live from your tracing data.
• Visualize the Blast Radius: RED metrics (Requests, Errors, Duration) are overlaid on every connection so you can see exactly where a failure starts and where it spreads.

Check rules can now query Dash0's own check metrics (dash0.check.*, dash0.issues.*, dash0.synthetic_check.*), enabling higher-level alerts based on the status of other checks.

Use this to:

• Aggregate check failures and group multiple related checks into a single alert, reducing alert fatigue by consolidating notifications.
• Monitor synthetic check execution and latency metrics: dash0.synthetic_check.runsCopy and dash0.synthetic_check.http.timing.*Copy

Find out more in Meta Check Rules

We've expanded the sorting options available in dashboard panels. You can now sort chart data by the Query name, giving you a new way to organize panels that contain multiple queries. This works across all chart types that support sorting, including Time Series Tables, Pie Charts, Hierarchy Maps, and more.

For Time Series Tables, the "Auto" sort option now sorts by value, surfacing your most significant data points first.

We've also tidied up the sorting dropdown itself: options like Value and Query name are now properly capitalized, and the Hierarchy Map's "sort groups by" field now shows inline help text instead of a tooltip.

When configuring charts that support the calculation field, you can hover over any calculation type in the dropdown to see a visual example of how it processes your data points within the query window.

Each visualization clearly illustrates the behavior of the calculation, including how it treats numeric values, Not-a-Number (NaN) entries, and edge cases. Whether you're deciding between "First value" and "First non-null value", or wondering how "Instant" differs from "Last", these guides will help you pick which one to use.

You can access these examples in Dash0 by hovering over the info icon next to the Calculation dropdown.

The images are static, it does not adjust to the reality of the data. But let us know if you find this kind of visual hints useful!.

You can now send telemetry data through Amazon's private network backbone without exposing traffic to the public internet.

Key benefits:

• Enhanced security with private connectivity
• Reduced network egress costs by keeping traffic within AWS
• Compliance with strict data security requirements
• Support for isolated VPC workloads

Available regions: eu-west-1 (Ireland), us-west-2 (Oregon)

For more information see https://www.dash0.com/hub/integrations/int_aws_privatelink/overview

Starting with version 0.92.1 of the Dash0 operator, Kubernetes events are now automatically collected in monitored namespaces.

The collected events will be presented in our brand new Kubernetes event view in Dash0.

You can also filter for Kubernetes events in the logging page by using otel.event.name=k8s.event filterCopy, or filter for specific events via the otel.event.nameCopy attribute.

See how Kubernetes events are emitted, collected by the Dash0 operator, and displayed in the new Kubernetes Events view: