Ntperf/ntperf64 probe can only be used to monitor Windows performance counters, therefore it is not suitable for this task.
Below instructions are based on an approach of monitoring local event logs.
Ntevl probe can be deployed to robot on the server where there are data collector set instances need monitoring.
When configuring ntevl probe, log file 'Microsoft-Windows-Diagnosis-PLA/Operational' should be added to the list of 'log files to be monitored' in the Setup>Properties tab of ntevl probe config window. This log file contains log events for Performance Logs and Alert.
Typical event ids that represent an error from data collector set are 1004 and 1016. Also contact local system administrator for an entire list of event id to be monitored from log 'Microsoft-Windows-Diagnosis-PLA/Operational'.
If the data collector set is assigned with a schedule, also monitor log 'Microsoft-Windows-TaskScheduler/Operational' for event ids for schedule failure.
Guides for how to create profiles/exclude rules in ntevl probe:
Ntevl Configuration Guide