Once the system based on the agent definition is registered in AIMS, its time to create the node that contains the properties, stattypes and events we want to monitor. If you try to click the system in the UI before this is done, you will see that there is no actual information ("no nodes") connected yet. The concept is that you can register multiple nodes of the system type within the same system if necessary, it all depends on the structure of what you monitor. In most cases a system will represent a unique application / system / server etc. So for instance for a server, it would be natural to create a single node that contains all the stattypes and events necessary for the monitoring (cpu, memory, I/O, etc)