Interface should reload performance counters after failover and keep track of them.

The interface should be designed to reload counters both at startup and when becoming the active interface after a failover. The interface should realize the counter is not available and periodically attempt to reload (ideally with a message to the log), not just reload at failover. The reason is that if a service is stopped, the counter will “disappear” (windows unloads it). When the service restarted, the counter returns and if the interface is not looking for it or has dropped the reference, the associated tag is now flatlined. It seems logical that when a failover occurs, you would want and most likely need to “refresh” the counter references in most cases (a failover basically should only happen infrequently and likely when other servers are impacted). However, currently the interface loads the counters exactly once, and only at startup, unless the /ReloadPerfData parameter is configured to reload periodically. Perhaps this logic change would allow users to leave it off avoiding any performance issues, but know that in the event of a failover, the counters will all be properly loaded.