Ingest /metrics data into SigNoz aka OpenTelemetry

Question

I use podinfo as a demo pod and want to ingest the exposed /metrics endpoint data into SigNoz. The SigNoz documentation describes a possible ingestion using Prometheus, so I could spin up a Prometheus instance to scrape the endpoint and then ingest into SigNoz.

Now I ask myself if I can do this directly with SigNoz without adding Prometheus? SigNoz states it uses all provided OpenTelemetry receivers, which is not helping, am I right? As I got this, OpenTelemetry is a tracing standard and not usable for metrics, wrong or write? Do you have any other suggestions to add metrics to SigNoz?

Answer 1

You don't need to run prometheus to scrape and then send metrics to SigNoz. OpenTelemetry collector has prometheus receiver which scrapes metrics the same way as does prometheus. You just need to add targets in the config of otel-collector-metrics service in SigNoz just like prometheus and that's all.

The docs link you shared shows the same thing. Just add the endpoint where metrics is available without the /metrics path in the below config as a new job_name.

  # Data sources: metrics
  prometheus:
    config:
      scrape_configs:
        - job_name: "otel-collector"
          scrape_interval: 30s
          static_configs:
            - targets: ["otel-collector:8889"]
        - job_name: "my-new-job"
          scrape_interval: 30s
          static_configs:
            - targets: ["localhost:8080"]