Batch scenario where a scheduled process runs at frequent intervals to keep other systems up to date. The requirement is not immediate replication, but close to real-time, from hourly down to every minute.
Messages are small and delivery is usually required to avoid systems from getting de-synchronized.
Even if guaranteed delivery is built into the interface, there is usually a full refresh weekly or monthly which is done via (SC-BDR), or as a high volume burst of individual messages or small batches, (SC-EDR or SC-SDR), in which case, this creates peaks of high-volume flows of data. These should be scheduled preferably out of business hours.
Both messages and files can be used, with messages being preferred. The only reason to use files is to re-use full-refresh extract/load scripts.
Transport of files is currently often FTP or FTPS, and for messages, there is local Message Queue (JMS, AMQP) servers.
As with scenario SC-EDR, there are often several subscribers, so the same choices should be made as for SC-EDR.
The use of ESB will drive re-use of the same data messages, and the publisher should only have to know what and when to publish something, while the ESB manages the further distribution of data.