what's the result of a critical time series query while the realtime data aggregation is not ready?

Hi guys,I’m wondering a critical case:
One hand, the realtime event data is continuously writing into the system with a certain timestamp such as ‘2016-05-01T00:00:00.000’.

Another hand, here comes a query with intervals like [ “2016-04-28T00:00:00.000/2016-05-01T00:00:00.000” ]

My question is that druid will return an incomplete aggregation result at the last time point?Or just pick out the last datapoint out of the result set?

Druid will return a result that reflect all data that has been loaded into the system so far. So, aggregates for the most recent time range will be partial.