Prerequisites
- Install and configure Nebius AI Cloud CLI.
- If you don’t have a service account for observability services, create one.
-
Make sure that the service account is in a group that has at least the
viewerrole within your tenant; for example, the defaultviewersgroup. You can check this in the Administration → IAM section of the web console. If the service account is not in the required group, click→ Add to group, and select
viewers. -
Issue a static key for the service account using the following command:
Copy the value of the static key from the
tokenparameter of the response. You will need it on later steps.
Enable trace collection
To collect traces from your Managed Service for Kubernetes cluster, enable traces in the Nebius Observability Agent for Kubernetes:-
Create or update your
values.yamlfile: -
Install or upgrade the Nebius Observability Agent for Kubernetes:
Send traces from your application
When traces are enabled, the Nebius Observability Agent for Kubernetes exposes an OTLP endpoint for receiving traces:<namespace> is the namespace where the agent is installed (default: observability).
Configure your application
Configure your application to send traces to the OTLP endpoint using OpenTelemetry SDK and OTLP exporter for your programming language: You can also find instrumentation libraries for your framework in the OpenTelemetry Registry. For applications that support OpenTelemetry auto-instrumentation, you can configure the OTLP endpoint using environment variables in your Managed Service for Kubernetes deployment:<namespace> with the namespace where the Nebius Observability Agent for Kubernetes is installed.
View traces
After configuring trace collection, you can view the traces in Grafana. For instructions on how to connect and explore traces, see How to view traces in Grafana®.See also
- How to view traces in Grafana®
- Nebius Observability Agent for Kubernetes
- OpenTelemetry documentation
- OpenTelemetry Registry: find instrumentation libraries for your language or framework