Analytics offload issues on Kubernetes

If your Analytics offload fails, you can perform checks to help determine the cause. Common causes for an offload failure are connection or security-related issues.

When you save the configuration files, their syntax is first validated to verify that it can be applied. However, this does not confirm that other parts of your analytics collection ecosystem are trouble-free, for example, that your endpoints are running, or that the TLS certificates are validated. Errors in these areas may also result in no data output.

To verify issues you will need to examine the logs for one of your Kubernetes analytics ingestion pods. (Note that it may take a few minutes for the pods to pick up changes in the configuration files after edits are applied.)

For information on obtaining logs, refer to Configuring remote logging for a VMware deployment and Configuring logging for a Kubernetes deployment.

Following are some common checks you can make to troubleshoot the offload:
  • Verify that the specified URL or the host and port values for the target system are accurate.
  • Verify that the server or cluster of servers in the target system are running and reachable. If applicable, also verify that the connection problems are not caused by firewall settings.
  • If you are using Transport Layer Security (TLS) to establish a private and secure communication channel to the target system, investigate whether the issue is caused by SSL certificate errors; for example, untrusted server certificates, certificates that have expired or are not yet valid, or missing intermediate (or chain) certificates. For more information on using TLS, see Creating a TLS profile.
  • If offloading to an Elasticsearch or Kafka target, verify that your configuration settings are valid; for example, the user credentials for authenticating to Elasticsearch or the Kafka topic name.