memory consumption, performance, reliability

[kagahd]

I don't know nodejs but is it true that you're loading the source collection first in an array to insert then each document one after another to the target collection? See:

mongo-clone/src/index.js

Line 36 in 0f904a2

const allData = await sourceCollection.find().toArray();

How about memory consumption for big collections? How about bulk inserts in batches? Is your tool able to keep the target collection in sync once cloned? If not, why not just using mongodump/mongorestore which also restores indexes automatically?