It is very common to use Secondary nodes to do some analytics jobs, and it is working perfectly with common query, since it is possible to use a readPreference (from the client or during a call)

However the MongClient, MongoDatabase, and MongoCollection force a ReadPreference.primary() when listing schema information.

This is blocking when the applications only have access to secondary nodes in a replicaset.

We should add the ReadPreference as an optional parameter to schema read/list methods