To be able to monitor interfaces of an on-premise system you first have to add it to Interface & Cloud Monitoring and select the monitoring categories you want to monitor for the system. 

1. Open the Advanced Integration Monitoring Application by clicking on the 'Integration & Cloud Monitoring' tile in transaction FRUN
2. Click on the pinwheel in the upper right corner of the 'Integration & Cloud Monitoring' application
3. In the configuration panel expand the 'Technical Systems' tray and click on the pen icon in the upper right corner. This will open the 'Integration and Cloud Monitoring - Systems' view. 
4. If your system is not on the list yet, click the 'Add' button to add it. If it is on the list, click on the system ID of the system. 

In the next step, you see all monitoring categories which are available for the system, depending on the system type and the installed software components. Some recommended standard monitoring categories are preselected.

Select the monitoring categories in scope.

After selecting your monitoring categories you have to maintain filters to define what exactly you want to monitor. Some monitoring categories come with standard filters that usually just select all items of this monitoring category. SAP Focused Run can handle this high amount of monitoring data, so you can stick to this standard filters. Or you can set up filters of your own. 

Please note that the filters you set in the 'Monitoring' step filter the data which is collected from the managed system or cloud service. You can also set up filters later in the 'Alerting' step, which will filter the data already in Focused Run.

You can create more than one filter for a monitoring category.

1. Select the view 'Monitoring'
2. To add a new filter click the '+' button on the upper right corner of the table
3. Select the Category for the filter. The drop down list will contain all activated categories for the system
4. Enter a filter name
5. Select the target client for the ABAP system / the end-point for the cloud services

The setup of the filters for the monitoring categories only makes sure that the data is collected, however, alerts are not created automatically. To create alerts and notifications you have to create an alert individually. 

1. Go the the view 'Alerting'
2. Click on the '+' button in the upper right corner of the alerts table. A new panel will appear below the table for you to enter the alert information
3. Select the monitoring category
4. Select the metric name. The available metrics depend on the monitoring category.
5. Enter a name for the alert

In the next sub-step, you have to maintain the filter. For most monitoring categories the available filter fields will be the same as for the 'Monitoring' configuration described above. 

For some metrics, you have additional filter fields or you have to maintain metric parameters. You can check the collected data in Focused Run to determine which filter values to use for alerting. Most of the fields can be found in the Collection Context of the collected data.

Please note that the filters on 'Alerting' level filter the data that is already available in Focused Run. They do not influence the data collection itself. You can only filter for data that is collected. Be aware of this if you created strict filters in the 'Monitoring' step.

1. Enter a name for the filter
2. Maintain the necessary filter fields

In the last sub-step you have to activate the alert:

1. Check the check box next to 'Active'
2. You can adjust the calculation frequency and the severity
3. Available threshold types depend on the metric selected. If the threshold type allows it you can adjust the threshold value which triggers the alert.
4. Select the notification variant and the outbound connector variant from the drop-down list.

For single exceptions, the threshold type is always 'Already Rated'. This means depending on the calculation frequency the number of exceptions is checked and an alert is created if this number is bigger than 0. If you want to reduce the number of alert for these metrics, you could increase the value for the calculation frequency to increase the time between checks.