I have data in mongodb and I want to use some of the properties saved in that database to start and maintain a graph database using amazon neptune. I can use aws dms to get the initial batch of data, but how about maintaining and updating the data as it changes in mongodb? Essentially I need to keep the neptune database synced up with the mongodb database because mongodb remains my primary database. As a simple example, consider that mongodb contains, among other things, users' lists of subscribers. I want this info available for graph queries in neptune, but changes to that info are saved in mongodb primiarily so they need to be forwarded to neptune to keep neptune current. Is there a best-practice for doing something like this? The bulk loader and dms options implicitly assume that I'd be inserting large amounts of data at a time, but these may be simple small changes like one user subscribing to another user. What's the best/correct/recommended way to go about this?