Modeling how the infrastructure monitoring tools are working is a little different than normal components. If iiWS is down and no events are going though, it does not mean that anything is harmed other than iiWS. iiWS could simply have been processing OK messages. It also does not hurt the cell. The cell is still processing along.
The way I have seen it done is to have a higher level logical component such as SAP monitoring and have sub components of all the programs used to monitor SAP. Then when one of these has a problem the logical component will show their is a problem.
You can also have logic groups of all the cells, all the adapters, and others like that.
This approach gets you out of trying to build a topology model.
Hope this helps,
Apparently you know when/how SIM, cells, etc environment, is impacted and in which degree, so I think what you need to do are: streamline the logics of the situations, locate the impact driving events (you may need to refine the events or even create some mechansims to generate them), and associate the driving events with the components (SIM, cells, adapters, etc). This way, when the identified events show up, the components in your model will be reflecting.
Thanks for the thoughts. I guess at the end of the day, one thing does 'impact' the other, so it could be set to flow an 'impacted' state via a 'decreasing status' up through the model. Or possibly could try and get some logical containers in there representing event flows (as a business process icon??), which could chain into a 'Event Management Flow' 'service' or something like that??
The fact you are asking these question and the way you explain them suggests to me you already have a pretty good idea , its a huge subject this and there are very few sources of usefulproduct specific documentation out there, I have attached one which is an excellent source of reference and I am sure you will find it useful....