Benefits

The load balancing exporter can send telemetry data to multiple back ends at once using a routing policy. You can use routing policies to sort telemetry data into groups and map these groups to specific endpoints.

Using the load balancing exporter, you can also send data to other running instances of the OpenTelemetry Collector through Collector endpoints. For example, you can send all traces to one running instance of the Collector and all logs to another running instance. Using this method, you can process or manipulate your data in separate Collector environments.

Get started

To use the load balancing exporter, provide a list of back ends that can receive telemetry data.

You can provide a static list of IP addresses, or a DNS host name to resolve. See Sample configurations for example configurations of the load balancing exporter that use each type of list.

Routing keys

The load balancing exporter can export spans, metrics, and logs depending on the routing_key configured.

The routing_key groups spans, metrics, and logs together. For example, the load balancing exporter sends all spans that belong to the same traceID to the same back end.

The following table shows options for the routing_key:

By default, the routing mechanism is traceID for traces and service for metrics.

Sample configurations

This section details some example configurations for the load balancing exporter.

Static list

The following example uses a static list of host names to configure the load balancing exporter for separate back ends:

receivers:
  otlp:
    protocols:
      grpc:
        endpoint: localhost:4317

processors:

exporters:
  loadbalancing:
    routing_key: "service"
    protocol:
      otlp:
        # all options from the OTLP exporter are supported
        # except the endpoint
        timeout: 1s
    resolver:
      static:
        hostnames:
          - backend-1:4317
          - backend-2:4317
          - backend-3:4317
          - backend-4:4317
      # Notice to config a headless service DNS in Kubernetes
      # dns:
      #  hostname: otelcol-headless.observability.svc.cluster.local

service:
  pipelines:
    traces:
      receivers:
        - otlp
      processors: []
      exporters:
        - loadbalancing
    logs:
      receivers:
        - otlp
      processors: []
      exporters:
        - loadbalancing

Kubernetes resolver

The following example configures the load balancing exporter for a Kubernetes resolver:

receivers:
  otlp:
    protocols:
  grpc:
    endpoint: localhost:4317

processors:

exporters:
  loadbalancing:
    routing_key: "service"
    protocol:
    otlp:
        # all options from the OTLP exporter are supported
        # except the endpoint
        timeout: 1s
    resolver:
    # use k8s service resolver, if collector runs in kubernetes environment
    k8s:
        service: lb-svc.kube-public
        ports:
        - 15317
        - 16317

service:
  pipelines:
    traces:
    receivers:
        - otlp
    processors: []
    exporters:
        - loadbalancing
    logs:
    receivers:
        - otlp
    processors: []
    exporters:
        - loadbalancing

Settings

The following table shows the configuration options for the load balancing exporter:

Troubleshooting