1

I'm using Datadog helm chart. I've got the exact same installation in two different clusters, both clusters have the exact same configuration.

In one of them I'm collecting custom metrics, plus cluster metrics.

In the other one I'm attempting collection of custom metrics, plus cluster metrics, but cluster metrics are not being collected.... I don't understand why, and I don't even know where to start diagnosing it.

caeus
  • 3,084
  • 1
  • 22
  • 36

1 Answers1

2

GKE autopilot does not collect logs for datadog, datadog itself collects, monitors and pushes to datadog.

Not sure how you installed with helm as it was not share but there is something called providers with datadog.

Provider : providers.gke.autopilot

helm install <RELEASE_NAME> \
    --set datadog.apiKey=<DATADOG_API_KEY> \
    --set datadog.appKey=<DATADOG_APP_KEY> \
    --set clusterAgent.enabled=true \
    --set clusterAgent.metricsProvider.enabled=true \
    --set providers.gke.autopilot=true \
    --set datadog.logs.enabled=true 
    datadog/datadog

There could be other things to check: is Agent running, is Daemonset running?. Autopilot mostly sets(round off) the resources if anything wrong.

caeus
  • 3,084
  • 1
  • 22
  • 36
Harsh Manvar
  • 27,020
  • 6
  • 48
  • 102
  • I have two clusters, both are gke autopilot. Both have exactly the same configuration. Both have providers.gke.autopilot set on true. I used helmfile for the installation. DaemonSets are up and running correctly. Logs are being collected in both clusters correctly. Custom metrics are being collected in both clusters correctly. K8s metrics (ie kubernetes.containers.restarts) – caeus Jul 06 '23 at 16:17
  • that's weird anything in logs that's helpful or in events hope there is no error in one failing. – Harsh Manvar Jul 07 '23 at 05:19