"Perfomance-triggered" automatic E2E Trace Recording (HTTP Script only)
Normally, it is not useful to run a EEMon Script with an increased trace-level to avoid performace issues on the related backend systems. Therefore the standard tracelevel is set to "low".
If you need an E2E Trace to analyse a exceptional script behavior you can trigger a trace run with higher trace-level from EEMon Real Time Monitoring I at any time.
Sometimes it is useful to do this automatically:
- Navigate to "Solution Manager Configuration" workcenter in SAP Solution Manager
- Select "Technical Monitoring" in left menu
- Select "EEMon" radio button
- Navigate to "Monitoring" step in guided prodedure
- Select the scope Script
Set the value for parameter "schedule.min.autotrace.period" to 3600. This means once per hour an additional script execution might be triggered. It is not recommended to set this parameter too low (e.g. to 10) otherwise with each script execution tracing is anebled in the managed system and each script execution might trigger an additional execution.
Note that - as opposed to other configuration parameters - the parameter schedule.min.autotrace.period can be set only on scope Script. Setting it on scope Global or Script on Robot will not have any effect. The normal hierarchy of configuration scopes is not effective here for performance reasons and to minimize risk for monitored systems due to accidental use.
Please also note that this parameter has no effect on SAGUI script.
- EEMon will triger an extra execution for a http EEMon Script which showed a bad performance during its regular run.
- EEMon will use the trace-level defined in temporary configuration
- EEMon will not extra-execute the script if it has already been extra-executed within the value of schedule.min.autotrace.period (seconds) to protect a system against a "analytical DoS".
To ensure that the extra execution is executed immediately and not serialized with other script executions you can set the property "semaphore.use.default" in temporary scope to "false.
You can do this in the monitoring UI -> Administration -> Temp Configuration -> add
If the script is scheduled to run every 10 minutes, always gets red performance rating, and you set the property
schedule.min.autotrace.period=3600 then one additional execution with trace is allowed per hour.