I'm reading through Jech's book, and in the section Stationary Sets in Generic Extensions (pages 444 and 445), he remarks that the poset used in shooting a club through a stationary $S\subseteq\omega_1$ can be generalized for regular $\kappa>\aleph_1$, but with additional restrictions on the stationary subset we use in order to preserve $\kappa$.
Why do we need these restrictions? What goes wrong in the usual argument if we consider, say, $\aleph_2$?
Is it that all club subsets of ordinals $\alpha<\omega_1$ have countable cofinality, while those $\alpha<\omega_2$ might have confinality $\omega_1$ or $\omega$?