I've had conflicting info from some of our engineers about this. Some say the nodes are added to the Node State and Interface State alerts; I just tested it with an ICMP only and an SNMP node and both times they were not added to the alerts.
Look at each of your node object alerts and check the trigger condition statements declared in it.
If you'll add new nodes, keep a tab in your mind that you have to make sure that that new node will be within the alerting scope of the node object alert you have.
Say you have a SW alert that fires when CPU load is greater than 95%, check if that SW alert has other lines in its trigger condition that checks further for other criteria say via node custom properties. Maybe that sample SW alert also checks for a 'Environment_Class' node custom property, that that alert will only fire for nodes assigned with the value Environment_Class = Production.
So if you've added a new node, even if it is indeed in the Production environment, but you've forgotten to populate the Environment_Class node custom property, or may have keyed in something else in there other than the word 'Production', then that newly added node will definitely lie outside the scope of the cpu alert described above, and will definitely not fire an alert action even if that node's CPU exceeds 95%.
I hope this helps.
