1. They are local/relative extrema by nature but whether or not they are absolute/global extrema depends on the interval. Let's say an upward cusp (i.e. a local/relative maximum) occurs at x = a. There could possibly be some x = b such that f(b) > f(a) on your interval in which case, your cusp is not a absolute/global maximum.
2. Depends. If the two relative minima are the absolute minima, then there you go, there's your absolute minima. However, if not, then there's always the possibility of having endpoints that are lower than your minima but are not defined. Then you don't have an absolute minimum at all.
