SQL Failover Cluster Monitoring Best Practices

I am trying to best understand what the best practices are for monitoring a SQL Failover Cluster.  My scenario is as follows:

SQL Server 1

SQL Server 2

SQL VIP

I add both the actual SQL nodes into Orion as well as the SQL VIP.  All of the resources that follow the VIP are monitored on the VIP node.  The part I am struggling with is how best to do the application layer monitoring.  The SQL Performance monitoring via AppInsight is typically applied on the SQL VIP so it follows the active node.

Where is the best place to put the Microsoft Windows Failover Cluster app monitor, on the SQL Servers or on the VIP?

Thanks in advance for any thoughts on this!