So I currently have Chef code (Windows and Linux) that installs the SolarWinds agent and registers the nodes with the Orion server. Right now I've been testing this in the dev environment, but that has a lot of churn. I'm not sure the best way to delete the node from Orion when the VM is being shut down (adding an API delete step?), or if I should be looking at moving to a different model like node discovery... would that properly handle marking nodes as down and then reconfiguring the name/configuration if the IP gets recycled?
This is less of an issue for prod as we don't cycle servers very often there, and it's at predictable times so we can easily add a step to decommission them from Orion as part of the procedure. But dev does have a lot of churn, and we want to be able to test out our monitoring configurations there before prod... without being unnecessarily spammed about nodes going offline because the VMs were terminated. And the devs don't have access to remove nodes from Orion, either, so I can't ask them to clean things up as part of their process.