HashMap values always got overwritten by the last input key in Hadoop MapReduce

Question

I created a hashmap in the reducer step. The purpose is to collect all results from reducer and then further process it in the cleanup() method.

My Hashmap has computed <