There are some mongodb data that need to be statistically analyzed, and there are correlations among multiple collections.
If mapReduce is used, it will be too complicated and the performance will be very poor.
What is the usual solution?