I've been seeing a ton of false positive alerts for servers in our infrastructure. These alerts are mainly Physical CPU & Memory. I've tuned the alerts to only fire off if the condition exists for 1 minute, and that cut the number of alerts in half. I'm still seeing the alerts, mainly from about 10 servers. I've used Spotlight on Windows to real time monitor CPU and memory on those servers and when an alert is fired off, there's about a 50% chance that it's a legit alert. The alerts are the out of the box alerts, with no special changes. I'm also polling via WMI. The servers in question seem to be for the most part VM's also.
Am I missing something? I can't be the only person to have had this issue before.