I've been running an enterprise application in GlassFish v2-b25 for
about 5 days. Today I come back in and test and the java process
running glassfish is using 98%+ of the available CPU and the logs
don't say anything about errors (exceptions, out of memory, etc.). I
can log into the GlassFish Administration application fine but not
the web-tier of the application deployed. Has anyone seen this
behavior or know what could be going on?
The Web-tier application is Openlaszlo based (can sometimes need lots
of file descriptors, but I am not getting those errors).
Enterprise tier is getting a connection from a connection pool.
Everything is Java EE 5 based except a few web services.
Any additional profiling or telemetry I can turn on to better
determine what is going on (increasing the verbosity levels)?
Thanks,
Iran.