"typical" is hard to define, since it very much depends on the hardware, number of nodes run, datastore type and the shape of the data itself