In my use case,I deal hundreds of millions of data once an hour using hadoop mapreduce program.Every record has to query on latitude and longitude to find out nearest place.And I have about 67millions geo point stored in aerospike.But performance seems insufficient.statemenet like this stmt.setFilters(Filter.geoWithinRadius(locbin, lng, lat, radius)); stmt.setAggregateFunction(“geo_filter”,“geo_nearest”, Value.get(lng), Value.get(lat)); I can’t think of ways to improve，any one can help？