How to restrict mongo-spark-connector making too many connections to mongo cluster?

Question

I\'ve built a simple data exporter job in spark using mongo-spark-connector. This job reads collections from mongo and writes it into S3. However, for collections of huge size (