I would like to know: is it is possible to integrate Apache Hadoop or MapReduceFramework on a grid computing environment?
Certainly it is possible, I have seen it in play.
IBM does it with its Spectrum Symphony grid middleware platform.
For details on the solution read here: https://www.ibm.com/support/knowledgecenter/en/SSZUMP_7.1.2/sym_kc/sym_kc_managing_mapreduce_framework.html