i think starting point should be how much data do you have. We work in a similar area and have found custom parallel solutions much more efficient than Hadoop or Hadoop-like solutions on Cloud Platform. We have our own tool "SPHeRe", that utilizes ontologies for semantic heterogeneity resolution over cloud platform in parallel.

I would suggest to test cloud vendors also, create accounts on their platforms and test their interoperability issues.

Data quantity is not a problem. It can be easily downloaded from the net and converted into the desired source formats.

@Muhammad Amin & @Lucian Hancu :

Thank you for responding, i want perform semantic interoperability analysis on various EHRs source, which is from heterogeneous cloud resources.

I`m stuck in data classification of various EHRs data.i`m unable to find clear cut idea about how i perform semantic interoperability test on those EHRs datasets.