The IETF has standardized about 100 MIB modules. These data models are very detailed and concrete and new people usually have problems to find and understand the core pieces of these data models. In fact, some people claim that the IETF lacks a documentation of the information model which is the conceptual basis behind these MIB modules. (For more information about the distinction between data models and information models, see A. Pras and J. Schönwälder: On the Difference between Information Models and Data Models, University of Twente, University of Osnabrueck, RFC 3444, January 2003.)
Information models can be specified at different abstraction levels. On a very high abstraction level, an information model just represents very abstract concepts (such as interfaces, physical components) and how they relate to each other. At a more detailed abstraction level, an information model will represent attributes and how certain relationships are modelled, closer to the real data models.
Sharon Chisholm started to put together a very high-level information model for some of the core Internet MIB modules. The UML diagrams can be found here:
Jürgen Schönwälder and Andreas Müller have developed an algorithm to reverse engineer concrete information models from MIB modules. (For more information, see J. Schönwälder and A. Müller: Reverse Engineering Internet MIBs, in Proc. 7th IFIP/IEEE International Symposium on Integrated Network Management, Seattle, May 2001.) Of course, the UML diagrams produced by the implementation of this algorithm need some fine tuning in order to make them readable. The following collection of revised UML diagrams is currently available:
This information is provided as is. But still, we like to get feedback so that we can improve and extend the models. So if you have any comments, please send them to the NMRG mailing list.