Does all definitions need a reference? Is there any rules to when you need one(or more) and when you don't?
You need to make crystal clear what is your contribution and what isn't. In general, all definitions that are not your contribution need the proper reference.
In practice that could get cumbersome, so you end up not citing what is considered common, basic knowledge. However, that line is somewhat blurry and one needs experience to do it right.
You can check good papers in your field to guess what is/isn't common knowledge. Another interesting tactic, that I often use, is to say:
"Since we are using XXX, we will briefly review the involved concepts [X]".
That way it is clear that they are not your contribution and that reference [x] can be consulted for further details, without too many references and unnecessary text...
To add clarity, I am assuming there have a set of terms you wish to include in a publication (be it conference, journal, or thesis), but confused whether or not to cite the definitions for which.
You may cite a term if it isn't too common in your research topic.
If you are pretty much sure that the term you included do not stray far from the common sense understanding within the domain in which your research concerns, you do not need to cite any relevant material for that term.