Hi,
In the Denodo platform, even complex distributed queries that need to process billions of rows can be usually resolved by moving very little data through the network, provided the right optimization techniques are used. The Denodo optimizer uses a variety of techniques for this purpose including query delegation, query rewriting, advanced join optimization, ‘on-the-fly’ data movement and parallel query processing. Once you configure the Hadoop data source, you can refer to these documents below for best practices in [optimizing the queries](https://community.denodo.com/kb/Performance%20&%20Optimization).
Once you’ve reviewed the optimizations, and if it is necessary to do the processing in denodo, you can also configure the MPP acceleration, for larger datasets. You can find more details [here](https://community.denodo.com/kb/view/document/MPP%20Query%20Acceleration:%20Sizing%20guidelines?category=Performance+%26+Optimization)
Hope this helps!
