Intermittent Windchill Method server crash due to dynatrace

Platform Open Q&A
, ,
1

Has anyone encountered a similar issue where Dynatrace OneAgent instrumentation or automatic injection causes a java.lang.StackOverflowError in a PTC Windchill environment?

Based on our log analysis, repeated calls involving Dynatrace’s AgentInjectingWriter.flush() and Java’s PrintWriter.flush() appear to be contributing to the error. PTC has indicated that this error is triggering the Windchill Method Server to shut down.

We are currently using Dynatrace to monitor PTC Windchill. As an interim measure, we have disabled Full-Stack Monitoring/RUM in the affected environment to prevent further issues while the investigation is ongoing.

Could the Dynatrace community help clarify the following?

  1. What could be the potential root cause of the recursive or repeated calls between AgentInjectingWriter.flush() and PrintWriter.flush()?

  2. Could this behavior be related to a Dynatrace OneAgent instrumentation or injection issue, or a compatibility issue with the Java version or Windchill?

  3. Are there any known issues, configuration changes, or recommended fixes to prevent this StackOverflowError and the subsequent Method Server shutdown?

  4. What additional logs or diagnostic steps would you recommend to confirm the root cause?

Any insights or recommendations from anyone who has encountered a similar issue would be greatly appreciated.