Troubleshooting
If you observe unexpected behaviors or results, the following pages help you resolve issues. To diagnose the problems, enable logging and tracing, and use the Flink web interface. If no installation fault is the cause of your problem, consider how your installation operates.
Troubleshooting installation
- Kubernetes deployment
- You can check how the installation of the IBM Cloud Pak® for Automation platform rolled out by consulting the operator log files. See Troubleshooting operators.
- Single-server deployment
- If your IBM Business Automation Insights deployment on a server fails, see Troubleshooting the deployment of IBM Business Automation Insights on a server for causes and corrective actions.
Troubleshooting operation
- Common problems
- You observe that the expected events don't flow to Kafka or they do flow but do not display in dashboards.
- Causes
- The causes are most probably with Kafka, with the Elasticsearch pods or security policy, or with the Flink jobs.
- Solutions
- Learn how to log and trace, and how to use the Flink web interface to diagnose and fix errors.
- For a Kubernetes deployment, go to the troubleshooting page for your event emitter under Troubleshooting IBM Business Automation Insights on Kubernetes, or to one of the Troubleshooting Apache Flink jobs or Troubleshooting Elasticsearch and Setup pages.
- For a server deployment, go to Troubleshooting Apache Flink jobs or Troubleshooting Elasticsearch.