Post stratification question

I am planning to create post stratification weights for my dataset based on a few variables, so I just wanted to understand a few things:
-My sample is divided into sub-regions to produce estimates by each sub-region. I wanted to check if I should create post-stratified weights for each sub-region instead of entire sample. How is that going to affect my overall estimate (al sub-regions comibined)?
-I was thinking of using raking as a method for post-stratification. Or should I just keep it simple and multiply weights generated from each of the post-stratification variable?

Appreciate any advice on this.

thank you