If Rhapsody has crashed, start by noting the date and time when the crash occurred. Then review the following files for information that might explain the crash:
wrapper.log- check for log file for entries at about the same time as the crash for entries. Common issues include watchdog restarts of the Rhapsody process and JVM crashes. A watchdog restart of the Rhapsody engine may have been caused by resources starvation as described in Rhapsody Not Processing Messages.log.txt- check the log file for entries just before the crash for entries that might indicate what the Rhapsody engine was doing. For example, a message search or backup.- Check the top-level Rhapsody directory for any recently created Java crash reports (these files have filenames with the following pattern:
hs_err_pid****.log). - If Rhapsody automatically restarts, review the memory graph for last 1 hour and last 24 hours from the System Monitor page in the Management Console. Check for high memory usage or a saw-tooth pattern. This could indicate that the Rhapsody process requires more memory to process the message volume.
- Monitor the performance of the server (in particular CPU, Memory, and Disk I/O statistics) to determine whether they are high on the server.
Where the crash has been caused by the service wrapper automatically restarting the Rhapsody process on an irregular basis, check the hardware requirements. Rhapsody is sensitive to the hardware it is running on (especially in a Virtual Machine environment) and misconfigured virtual machines are a frequent cause of restarts.
More complex crash issues (especially where a crash dump has been generated) should be reported to Rhapsody Support.