OpenTelemetry internal telemetry Assets

This package contains dashboards that visualize the internal metrics from OpenTelemetry components.

The OTel collector dashboards are compatible with the metrics defined here in the OpenTelemetry collector documentation. The oldest tested version of the OpenTelemetry Collector in combination with this package is v1.44.0.

Furthermore, the package also contains dashboards to visualize metrics from the tailsamplingprocessor. Those metrics are documented here.

The dashboards query metrics from the collectortelemetry dataset. Set the resource attribute data_stream.dataset to collectortelemetry on the collector that emits internal telemetry (see below).

The dashboards rely on field names defined in above documentations.

You need to follow the documentation of the OpenTelemetry collector to setup and send internal telemetry to your cluster.

The most important prerequisite is to define the telemetry section under service and to set data_stream.dataset to collectortelemetry.

service:
  telemetry:
    resource:
      data_stream.dataset: 'collectortelemetry'
    metrics:
      readers:
        - periodic:
            exporter:
              otlp:
                protocol: http/protobuf
                endpoint: https://backend:4318

With this, you'll have internal telemetry on normal verbosity level.

The above configuration defines an OTLP endpoint that sends internal telemetry to a target collector over an OTLP connection. This target collector then exports the internal telemetry data to Elasticsearch using the elasticsearch exporter. The integration subsequently reads this internal telemetry data from Elasticsearch.

For all the other config, refer to the upstream documentation.

This integration includes one or more Kibana dashboards that visualizes the data collected by the integration. The screenshots below illustrate how the ingested data is displayed.