Troubleshooting

Reconnect TEA agent

Whenever the data displayed in Application Monitor is not in sync with the configurations provided or incorrect data is displayed for the field, you can use the Reconnect TEA Agent option. Reconnecting TEA agent creates new JMX connections to the entities and re-emits the status of all entities. For more information, see Reconnecting TEA Agent.

View Unreachable Nodes

While starting an Application Monitor TEA agent, all the AMXCE nodes are started. Sometimes, a JMX connection cannot be created for a specific AMXCE node. To view the nodes that have not started, click the Unreachable Node icon in the upper-right corner on the AMXCE Nodes page. The count of unreachable nodes is also displayed in the upper-right corner. The AMXCE node is unreachable only if the pod is running but the agent is unable to create a JMX connection. You can find pod id and namespace from the Unreachable Node tab. You can navigate to kubectl and check that pod status, and logs to troubleshoot the issue.

Entity Level Logs

To troubleshoot the entity-specific issues, you can use loggers. By downloading the entity-specific logs, you can debug the errors. By changing or adding different logs and levels, you can generate detailed log information for the entity. You can view or download logs for agents, AMXCE nodes, and applications. You can use Docker logs or kubectl logs along with Application Monitor logs. The Application Monitor uses logback for logger configuration. It has a default logback.xml file, located at amsgce-runtime-<version>/teaagent/config/logback.xml. To update the logging configuration, change this file before creating the TEA agent Image. After the image is built, you can mount the logging file but you will need to restart the TEA agent container.

Application Status

The Application Monitor displays the state of an application by aggregating and summarizing the state based on all its components, including the bindings. For example, If some application components are running and some are explicitly stopped, the runtime state shows Partially Running.

If an application is showing a different status than expected, check the status of different application entities such as components, services, references, and service or reference bindings. As the status of all child entities is included in the application status, you can identify the exact entity that caused the issue.

AMXCE Node Health Check

For troubleshooting issues related to the AMXCE node, you can use the AMXCE node Health Check feature. AMXCE Node Health Check is enabled by default. Health check enables you to view information about the threads that are used and JVM memory being used. For more information, see Viewing and Downloading Node Health Check Information.

OSGi Diagnostic Console

The OSGi diagnostic tool provides node-specific OSGi information that can be used for troubleshooting and diagnostic purposes without requiring special OSGi configuration and node restart. The tool provides a visual representation of the information normally provided by the OSGi console commands that require configuring OSGi console port and node restart. You can connect to OSGi command-line by using the default port 2222 which is configurable. For more information, see Viewing Node Specific OSGi Information.