In certain situations it becomes necessary to reimport data from the process engine. 

This is e.g. the case when we add new features (e.g. incident information) or new aggregations that require new data from the engine. 

In certain scenarios the engine does not have the data anymore (e.g. because of history cleanup).

From Optimize perspective I still would like to use all the new functionality - even for old data - so that I do not have to reimport from the engine where the data does not exist anymore.

Potential solutions include:

• copy of engine history (Relational)
• raw history from engine in elastic without aggregations (complete dump)