Based on my experience with UCMDB, I've faced challenges while discovering servers (nodes). We need to define the IP list of discovery in RTSM and run jobs. Instead, over the years, if we need to add some new assets, we have to add the new IP in RTSM and run jobs again. From the agent side, we need to provide the data flow probe IP. We can do this things as one way: we install the agent at the server (node) level (define the IP of the data flow probe) and after that, our data flow probe will capture the data from the agent and send it back to the UCMDB. That way, we can remove the one step of adding the IP in RTSM. When we work in an environment, the production environment continuously changes and at that time we should have knowledge about the new server (node). However, the auto-discovery part should automatically discover the new CI (node/servers) in an environment and inform the IT person.