Available Reports

Http Activity

The Http Activity report includes all instrumented methods of the instrumented Servlet classes.  The methods include service, init and destroy.  For the ‘service’ method (used to process an HTTP request) the report shows the HTTP method (GET / POST, …) the URL (excluding the query string).



Jdbc Activity

The Jdbc Activity report includes all instrumented methods of instrumented classes implementing the JDBC package.  This report includes the SQL string, if applicable.  When the SQL string is collected, it is obfuscated to prevent the collection of sensitive information.




Gc Analysis

The Gc Analysis report includes data about the garbage collection.





Rectangular Callout: The top panel display the response time (how much time was spent actually performing the GC event)Rectangular Callout: The center panel shows the percent of heap that was “in use” AFTER the GC event.  A ‘Saw Tooth” pattern is common for the generational collector.Rectangular Callout: The bottom panel shows the percentage of time spent in GC Events. Example, if the JVM does a GC event every second and it take 100 milliseconds to do the collection, that would be 10%.


GC time is a good indicator to the health of the JVM and the garbage collection system.


It is a ratio of the time spent doing garbage collection to the total processing time. For example, if the JVM performs a garbage collection once every second (1000 milliseconds) and it take 100 milliseconds to perform the collection, the result would be 10%.


For most applications,

Less than 7% is Green (healthy)

7% through 10 % is Yellow

Greater than 10% means the application response time is ‘probably’ degraded due to garbage collection.



Gc Regression

The Gc Regression helps identify memory leaks in the  Heap by projecting the heap usage into the future using a statistical linear regression curve.  A positive slope is an indication that the application ‘may’ have a memory leak. 






The JVM CPU shows how much of the total system CPU was consumed by an individual JVM process.  This helps us understand if the load is balanced between multiple JVMs and helps identify that some other process is competing with the application for resources.






Application errors, from instrumented methods, are shown in the Errors report.   The report includes the Date and Time of the error, the  Exception (type of error) and the Stack Trace of the current thread.





Long Running Transactions


The Long Running Transactions report shows any transaction (or method call ) during the specified time period that took longer than the user defined time to execute.


When you select the Long Running Transactions report, addition selection criterion is available on the selection page to allow you to define what a long running transaction is.





Method Hangs


The Method Hangs report shows any method that is considered “hung”.  A “hung” method it one that started during the selected period but has not, as yet, ended.


Our expectation is that this report is normally empty and any data here is an indication of a problem (the more data, the larger the problem).