Monitoring and Logging
20/06/2026
Monitoring of Graph Modeling may happen on two different layers, on the application server layer or the the underlying operating system layer. Alternatively, it is also recommended to ping API endpoints of Graph Modeling regularly to verify the availability of Graph Modeling.
Graph Modeling comes deployed on Tomcat which provides a JMX management interface described in the Tomcat JMX Guide. Use any JMX compatible console or monitoring agent with it.
As an alternative you may also deploy a monitoring webapp like psi Probe to watch system parameters from within the appserver container. Here you can find the latest release.
On OS layer the Graph Modeling system will seamlessly integrate into such monitoring systems used like Nagios or any commercial alternatives. Typically you will want to use plugins of that systems designed to monitor JEE compliant app servers like Tomcat. Or you can of course also implement monitoring based on system logfiles like for example Splunk does. Please Contact us for further consulting on this matter.
The minimum level of monitoring we recommend is a basic health check by regularly pinging any Graph Modeling API endpoint from another external system such as StatusCake. An unsuccessful response to such an API request can indicate the unavailability of the server. Commonly used API endpoints for this purpose are
http://$POOLPARTY_SERVER_NAME/PoolParty/apifor Graph Modelinghttp://$KEYCLOAK_SERVER_NAME/auth/for Keycloak
where $POOLPARTY_SERVER_NAME and $KEYCLOAK_SERVER_NAME are the domain names or host names of the PoolParty and Keycloak servers accessible by the monitoring system. In most cases, the expected status code of the response should be 200, indicating a functioning service.
PoolParty log files are placed per default in the following directory:
GNU/Linux:
/opt/poolparty/tomcat/logs
Note
PoolParty relies on Keycloak for user management. For more information on how to configure the logging for user management, refer to the Keycloak documentation.
Initially the log level is set to WARN, to avoid too verbose output of the system. Graph Modeling is using log4j as a logging framework, which allows high customization of log output. The configuration of the log output is specified in the log file located in the following directory (default installation path):
Linux:
/opt/poolparty/tomcat/lib
Log files keep growing over time and tend to waste disk space if left unmonitored and unmanaged. Please use the tools provided by your operating system, for example the logrotate deamon available on Debian distributions.