For the complete documentation index, see llms.txt. This page is also available as Markdown.

OpenTelemetry (OTel)

Nected services can export metrics and traces to an OpenTelemetry (OTel) Collector, letting you plug Nected into your existing observability stack instead of relying solely on built-in monitoring.

Prerequisites

  • A running OpenTelemetry Collector reachable from your Nected cluster, with its endpoint and protocol (http or grpc) noted.

  • Access to the nected-values.yaml file used for your Helm deployment.

  • Helm access to redeploy the Nected release.

Configuration

  1. Open the nected-values.yaml file used for your Nected deployment.

  2. Under each service you want to instrument, add the OTEL_* environment variables shown below.

  3. Replace the placeholder values with your OpenTelemetry Collector's configuration.

  4. Redeploy Nected for the changes to take effect (see Applying the configuration).

You can enable OTel for one, several, or all of the services below- each is configured independently.

Nalanda

nalanda:
  envVars:
    OTEL_ENABLED: "true"
    OTEL_EXPORTER_OTLP_PROTOCOL: "<<http or grpc>>"
    OTEL_EXPORTER_OTLP_ENDPOINT_TRACE: "http://<<otel-collector-host>>:<<port>>"
    OTEL_EXPORTER_OTLP_ENDPOINT_METRIC: "http://<<otel-collector-host>>:<<port>>"
    OTEL_EXPORTER_OTLP_HEADERS:
    OTEL_TRACES_SAMPLER_RATIO: "1.0"

Vidhaan Executer

Medha

Garuda (optional)

Configuration parameters

Parameter
Description

OTEL_ENABLED

Enables OpenTelemetry integration for the service. Set to "true".

OTEL_EXPORTER_OTLP_PROTOCOL

Protocol used to communicate with the OTel Collector. Supported values: http, grpc.

OTEL_EXPORTER_OTLP_ENDPOINT_TRACE

Endpoint of the OpenTelemetry Collector for traces, e.g. http://otel-collector.observability.svc:4318.

OTEL_EXPORTER_OTLP_ENDPOINT_METRIC

Endpoint of the OpenTelemetry Collector for metrics, e.g. http://otel-collector.observability.svc:4318.

OTEL_EXPORTER_OTLP_HEADERS

Optional headers sent with every request to the OTel Collector (e.g. for auth).

OTEL_TRACES_SAMPLER_RATIO

Fraction of traces to sample, from 0.0 to 1.0. "1.0" samples all traces.

ENABLE_PROMETHEUS

(vidhaan-executer only) Controls Prometheus metrics collection. Set to "false" if OTel is already handling metrics and traces, to avoid collecting them twice.

Applying the configuration

After updating nected-values.yaml, redeploy Nected with Helm:

Once the rollout completes, the configured services will start sending metrics and traces to your OpenTelemetry Collector.

Supported tools

Since Nected exports telemetry in the standard OTLP format, you can connect it to any OTLP-compatible observability tool via your OpenTelemetry Collector, including:

  • Grafana: open-source visualization platform, typically paired with Tempo (traces) and Mimir/Prometheus (metrics) for dashboards and alerting.

  • Dynatrace: full-stack observability platform with AI-assisted root-cause analysis across traces, metrics, and logs.

  • New Relic: cloud-based observability platform offering unified dashboards, alerting, and APM built on top of OTLP data.

  • Splunk: log and telemetry analytics platform, useful when you want traces and metrics correlated with existing log data in Splunk.

  • Datadog: monitoring and security platform with APM, infrastructure monitoring, and dashboards driven by OTLP ingestion.

The setup steps are the same regardless of which tool you use: point your OpenTelemetry Collector's exporter configuration at that tool's OTLP endpoint (each vendor documents this for their own product), and Nected's services will keep sending data to the Collector as configured above — no changes needed on the Nected side when you switch or add a backend.

Troubleshooting

  • No data arriving at the Collector: confirm OTEL_EXPORTER_OTLP_ENDPOINT_TRACE and OTEL_EXPORTER_OTLP_ENDPOINT_METRIC are reachable from within the cluster (network policies and DNS are the usual culprits) and that the protocol matches what your Collector listens for.

  • Traces missing intermittently: check OTEL_TRACES_SAMPLER_RATIO; a value below 1.0 intentionally drops a portion of traces.

  • Duplicate metrics from vidhaan-executer: set ENABLE_PROMETHEUS: "false" if you've enabled OTel metrics for this service.

Last updated