ScrollAllRequest
Simple back pressure implementation that makes sure the minimum max concurrency between producer and consumer
is not amplified by the greedier of the two by more then a given back pressure factor
When set each scroll request will additionally wait on WaitAsync as well as
MaxDegreeOfParallelism if set. Not that the consumer has to call Release
on the same instance every time it is done.
The maximum degree of parallelism we should drain the sliced scroll, defaults to the value of Slices
Set a different routing field, has to have doc_values enabled
An optional search request that describes the search we want to scroll over.
Defaults to matchall on the index and type of T in the ScrollAllObserver<T>.
Note: both scroll and slice information WILL be overriden.