Typically, NetWeaver must be restarted to bring the system back online after a system failure. If NetWeaver was restarted without collecting heap and thread dumps first, valuable information pertaining to the state of the system during the outage is lost. In this case, it is still possible to obtain valuable information to investigate the root cause of failure.
This section describes steps to diagnose ME failure post-mortem. The same troubleshooting steps detailed here apply if the system recovers on its own without requiring a NetWeaver restart.
For ease of use in viewing an SAP Note, right click on the Note link and select Open in new window or Open in new tab.
When a system failure occurred in the past, you should:
- Run database diagnostic reports.
- Analyze automatically generated heap dumps.
- Analyze Wily and log files for the time in question to determine root cause.
Analyze Automatically Generated Heap Dumps
If a server node experiences an Out-of-Memory (OOM) event, a heap dump may be generated automatically by the system if the JVM is configured to do so.
Check the file system to see if a heap dump was automaticallly generated during the time of the system outage.
You will find heap dumps on the file system under /usr/sap/<SID>/<Instance>/j2ee/cluster/<server>
Use MAT to Analyze SAP ME Heap Dumps
Analyze Wily and Log Files
After you restart NetWeaver or after the system recovers on its own, you can use additional tools and examine system log files covering the time period when the problem was occurring to determine root cause of the issue:
Run Database Diagnostic Reports
Root cause of a system down event is often on the database side.
After you restart NetWeaver, run database diagnostic reports to analyze database activity covering the time period when the problem was occurring.
For instructions for running Oracle Database Diagnostics reports, see SAP Note 1457893- How to generate Oracle ADDM, ASH, and AWR reports.
SAP Note 1498757 - 'SAP ME 5.2 Survival Guide'