I wrote a blog post on this. We were using event sourcing and went with locking+batching method described. For other tables we wanted to read and move to a data

I wrote a blog post on this. We were using event sourcing and went with locking+batching method described. For other tables we wanted to read and move to a data warehouse we used the txid check. A bit hacky, but on AWS RDS we didn’t want to deal with wal intricacies. https://mattjames.dev/auto-increment-ids-are-not-strictly-monotonic/ https://mattjames.dev/auto-increment-ids-are-not-strictly-mo...