In MPP architecture(distributed cluster in distributed mode), on basis of what logic the data is distributed among the different servers?
I believe this is done via partitions and replication factors; certain queries will allow you to run in distributed mode and would benefit greatly from MPP. You may find further information here:
1 Like