We would like to gather data from vSphere, metrics like number of servers, their names, traffic information (this to be defined later). What would be the best place to start and what steps do we need to accomplish this? Any brainstorming suggestions, SDKs etc. would be greatly appreciated.
I would start by taking a look at the vSphere Web Services API service and, more specifically, the HostSystem (ESXi hosts), VirtualMachine (VM), and PerformanceManager (source of metrics) managed objects.
SDK-wise, it really depends on your language of choice. The vSphere Management SDK includes several different languages including Java and .NET. There's also the open source "*vmomi" SDKs, for which there is Python (pyvmomi), Go (govmomi), and Ruby (rbvmomi).