Skip to main content

​
Collector manager not in Ready state

If the collector manager is not transitioning to a Ready state, follow these checks:
  • Confirm the integration appears on the Integrations page and validate the release.
  • Confirm the integration status is Active and not Attention Required or Delete In Progress.
  • Ensure the collector manager pod has DNS and internet access, as it communicates with the Plerion Control Plane over HTTPS.
Check the collector manager pod logs:
Example output:
Check the DNS pods:
Example output:
If DNS pods are in a Pending or CrashLoopBackOff state, resolve the DNS issue first. After DNS is healthy, restart the collector manager pod and check if it becomes Ready.

​
Pods crash loop with exec format error

If the collector manager pods restart repeatedly and the container logs show:
the node architecture does not match the image. Collector manager images published before arm64 support was added are amd64 only, so they cannot start on arm64 nodes such as AWS Graviton. The health probes fail at the same time because the container never starts, so treat this as one problem rather than three. Upgrade to the latest collector manager, which is published for both architectures.

​
Retrieve controller logs

If the issue persists, review the collector manager logs for more detail.
  1. List pods in the plerion-system namespace:
    Example output:
  2. Retrieve logs for a specific pod (replace with the pod name you want to inspect):
    The -f flag streams logs in real time.
By reviewing these logs, you can identify connectivity or configuration issues. For deeper diagnostics, see Enabling verbose logging.