Hello,
I have a bit of a weird issue that I wanted to go over. In my environment we monitor our servers using two distinct SCOM Management groups, the SCOM instance is SCOM 2012. The Management groups are both separated from the servers under management by a WAN connection. The SCOM agents report to two local Gateway servers, one gateway server per management group.
In general our configuration is stable and alerts seem to flow correctly into the management groups. The issue I have is that alerts based on monitors aren't getting resolved when the condition is resolved. As in the server that had the degraded health state has had the condition resolve itself without the alert being closed in my management group. Sometimes, the health state for the computer will still report degraded after the alert condition is fixed. Additionally both management groups are not consistent, I'll have fewer active alerts in one instance than the other.
My question is how does the agent go about informing the Management Group that an alert condition has been resolved? If an agent evaluates that an alert condition has been resolved will it verify that the management group, or in this case groups, has received the update? Will an agent send that updated health state condition more than once? Is it normal for some communications from agent to management group to fail? Is it wise to periodically reset health states on agents in the event some communication has failed?
Thank you in advance,
Jason
Hope that helps! Jason