We have received a number of inquiries from new MongoDB users regarding using our OLAP engine against MongoDB sharded clusters.

There is concern about our mechanism of aggregation to get data from Mongo, which is to use agg pipeline commands.

Are there special concerns that we need to be aware of in a sharded environment?

We are essentially going to rely on Mongo to handle the data sharding and query for the data from a primary server via Agg Pipeline. We are working from the assumption that Mongo is already able to handle grouping & sorting at individual sharded nodes.

Is this an acceptable assumption?