Under normal circumstances RockSolid performs silent monitoring at the instance level. All monitored data is centralised to the RockSolid application servers and RockSolid repository. Operational analysis and compliance checking processes run in real-time on the RockSolid repository data and generate alerts from the central application. No alerts are raised or recorded on the monitoring target SQL Server instances themselves.
However, in some situations it is useful to record an alert on the SQL Server instance so a custom process can detect and or invoke custom actions (before considering this a direction you should first check out the RockSolid Automation functionality as this provides such functionality). In these situations, for certain alerts, you can request RockSolid to additionally raise a SQL Server error log entry for the event within the target SQL Server instance.
Enabling this option with instruct RockSolid to raise blocking events as a SQL Server error log error on the target instance being monitored. If this option is disabled, blocking alerts are raised only by the normal centralised RockSolid event/service request process.
The "Blocking duration threshold (secs)" option specifies the minimum duration of a blocking condition before it is raised within the SQL Server error log. This setting does not affect how blocking conditions are raise as events within the RockSolid central application.
This option should only be enabled in special circumstances when integration of a third party process is required. Under normal circumstances blocking alerting should be raised from the central RockSolid application.