Currently looking at it as a way side-by-side a legacy application with a massive ancient on-prem database, business logic embedded in sprocs, a small set of tables pushing columns limits with col names like notes2-notes22 but missing notes21 for a reason lost to time.
Comments
We're thinking that AWS DMS > AWS MSK is an easy way to bring Kafka in, and then modern applications can deal with Kafka as their source of truth. Kickstarting the never ending effort to "kill legacy".