Query.iterator() should accept a param to specify the fetch size of entries. For indexless query, this would affect the underlying stream and for distributed lucene queries it would set the underlying DistributedIterator's fetchSize