Show TOC

Procedure documentationSpecifying Alert Thresholds


Thresholds for yellow and red alerts specify tolerances for the response times of EEMon script steps. You can specify the thresholds at various levels:

  • You can specify thresholds for all instances of an EEMon script at the same time.

  • You can specify thresholds for individual instances of an EEMon script which are run by an EEMon robot. These thresholds overwrite the setting for all instances of the EEMon script, at script level.

Example Example

The response times of a host in Frankfurt are regularly longer than those of hosts in Paris and New York. You raise the thresholds of the host in Frankfurt, to avoid unwanted alerts.

End of the example.

To avoid having to calculate meaningful threshold values for each instance and step of an EEMon script, manually, you can calculate them from historical mean values, automatically.


  • To get statistically-significant mean values, you have monitored the execution of EEMon scripts for several days in real-time monitoring.

  • You have administrator authorization.

  • You are currently performing the Alerting step of the guided procedure for setting up End User Experience Monitoring in the SAP Solution Manager: Configuration work center.


  1. Choose the Steps tab.

  2. Expand the hierarchy for the technical scenarios, EEMon scripts, and steps.

    The global threshold values specified for the steps of the relevant EEMon script are displayed in the columns for yellow (green to yellow) and red alerts (yellow to red).

    The response time threshold values are in milliseconds.

  3. Choose Edit.

  4. Specify the threshold values for yellow and red alerts in the Green to Yellow and Yellow to Red columns.

  5. Optional: Calculate threshold values for individual instances of a script:

    1. Choose Thresholds on Robots.

      • The instance-independent threshold values for yellow and red alerts appear in the G2Y Inherited and Y2R Inherited columns.

      • The column MAX shows the longest measured response time in milliseconds.

      • The column Mean shows the mean measured response time in milliseconds.

      • The Count column shows the number of executions of the instance of an EEMon script, on which determination of the longest and mean values (in milliseconds) as well as the standard deviation are based.

      • The DEV column shows the standard deviation. If the value is a multiple of the mean value, the quality of the script executions fluctuates considerably. We recommend that you identify the cause. In normal operation, the value is a fraction of the mean value and the quality of the script executions is consistently high.

      Note Note

      EEMon is based on the maximum and mean value calculation and the standard deviation is based only on script executions for which the relevant business scenario was available (alert status OK, Poor Performance, Critical Performance) but not on timeouts or executions for which the text checks failed.

      End of the note.
    2. Choose Historical Values.

    3. In the Threshold History [Days] field, specify the period (in days) on which to base the mean response times, and choose Enter.

      Maximum and mean threshold values and the standard deviation are recalculated for the specified period.

      Recommendation Recommendation

      Choose a period in which there were no irregularities or statistical anomalies.

      End of the recommendation.
    4. Select one or more steps of an EEMon script instance.

    5. Choose Adjust.

      A dialog box appears.

    6. Specify the allowed difference from the historical mean value for yellow and red alerts in one of the following ways:

      • As a percentage in the Factor for Average Value field

        Example Example

        The mean response time is 100 ms. The allowed difference is to be 20%. You enter a value of 1.2 in the Factor field. A threshold value of 120 ms is calculated.

        End of the example.
      • As a fluctuation margin for the historical mean value in the Factor for Standard Deviation + Mean Value field

      Example Example

      The historical mean in normal operation is 100 ms, and the standard deviation measured is 10 ms. You enter a value of 2 (sigma) in the Factor for Standard Deviation + Mean Value field. Therefore, the tolerance is two standard deviations (20 ms) and the threshold value for generating an alert is 120 ms (2 x 10 + 100). 95.4% of all normally distributed executions are statistically below the threshold value of 120 ms. At 3 sigma, 99.7% of all executions are statistically below the threshold value of 130 ms.

      End of the example.

      Note Note

      To show the distribution of the measured response times in relation to the historical mean value as a graphic, select a step and choose Display.

      End of the note.
    7. Choose OK.

      The system calculates the values for the allowed difference and copies them to the relevant fields.

  6. Save your entries.

Note Note

To calculate thresholds from individual rules, you can export the table of threshold definitions into a Microsoft Excel table and upload it again.

Expand the area Download and Upload Threshold Values.

End of the note.


End User Experience Monitoring shows the status of the EEMon script executions based on current threshold values.

Note Note

To ensure consistent reporting, the system does not adjust the status of old executions when the threshold values change.

End of the note.