The following workflow is becoming more and more common in streaming architectures:

If Dgraph is to be used as a primary data store in a production system, it must support logical replication / change data capture.
Is this currently possible? If not, is it planned for the future?
It's not possible right now but we can do something like this in the future. Some users have asked for subscriptions. This would be a similar feature I suppose.
We could potentially support this at Badger level first, and then expose it via Dgraph.
After chatting with @gitlw, it seems like this could be implemented relatively easily on top on our current plans for datacenter to datacenter replication, which already uses Kafka.
Adding it to our 1.X plans!
For an example of a distributed DB that's doing this in what I would argue is the Right Wayâ„¢, see CockroachDB https://www.cockroachlabs.com/docs/stable/change-data-capture.html
Github issues have been deprecated.
This issue has been moved to discuss. You can follow the conversation there and also subscribe to updates by changing your notification preferences.

Most helpful comment
After chatting with @gitlw, it seems like this could be implemented relatively easily on top on our current plans for datacenter to datacenter replication, which already uses Kafka.
Adding it to our 1.X plans!