grafana/k8s-monitoring-helm

EKS Fargate example failing

Opened this issue · 0 comments

skl commented

I'm getting a couple of errors after following the previously working EKS Fargate example:

Error: UPGRADE FAILED: execution error at (k8s-monitoring/templates/kubernetes-monitoring-telemetry.yaml:2:3): Invalid configuration for gathering pod logs! When using logs.pod_logs.gatherMethod: "volumes", Grafana Alloy for Logs must be a Daemonset. Otherwise, logs will be missing!
Please set:
logs:
  pod_logs:
    gatherMethod: api
  or
alloy-logs:
  controller:
    type: daemonset

and

Error: UPGRADE FAILED: execution error at (k8s-monitoring/templates/kubernetes-monitoring-telemetry.yaml:2:3): Invalid configuration for gathering pod logs! When using logs.pod_logs.gatherMethod: "volumes", Grafana Alloy for Logs should not utilize clustering. Otherwise, performance will suffer!
Please set:
alloy-logs:
  alloy:
    clustering:
      enabled: false

Following the advice above, I end up with:

cluster:
  name: eks-fargate-test

externalServices:
  prometheus:
    host: https://prometheus.example.com
    basicAuth:
      username: 12345
      password: "It's a secret to everyone"
  loki:
    host: https://loki.example.com
    basicAuth:
      username: 12345
      password: "It's a secret to everyone"

metrics:
  node-exporter:
    enabled: false

logs:
  pod_logs:
    gatherMethod: api

prometheus-node-exporter:
  enabled: false

alloy-logs:
  alloy:
-    clustering: {enabled: true}
+    clustering: {enabled: false}
    mounts:
      varlog: false
  controller:
    replicas: 2
-    type: daemonset
+    type: deployment

But then this fails I think due to: