The napshot you use needs to be take with the last version of your componant.
So your 6th step should be :
6)rerun snap job, change the master to the new snap in the audit job, rerun the audit job.
(rerun the discovery job isn't useful exepted if you change the discovery rule because the discovery wasn't good)
If the snapshot isn't effected by a new part or rule based on that part then I can rerun with the original, in any case I agree rerunning the snap is always good practice.
My point is that creating multiple template versions each time one needs to add or remove a part to/from the original is seems tedious. Perhaps I'm not explaining this very well (or I'm just wrong!)
Intuitively, one expects a change to the template to cascade down, i.e. if I make a template change that excludes 1 of 5 of my previously discovered servers then that ought to be reflected in the components discovered on the next run. This seems not to happen.
I'll put together a case scenario to better show what I mean.
i think this is noted in the docs. if you remove (or add maybe) objects in your CT it doesn't get reflected on servers that already have the template discovered. look in the components section of the opsmgr.pdf doc i think. you could add the parts you're adding or removing to the discovery portion of the template as well to help filter, but i agree this is a pain in the rear sometimes.
The napshot you use needs to be take with the last
version of your componant.
You're confusing a "snapshot" with a "napshot." A "napshot" is defined as:
Napshot: Someone's shot at taking a nap. For example, "I just took the red eye to work at this client, and it feels like 3am right now. I wonder what my odds for a napshot are today?"