While leading tech for a fintech org where we built out a relatively large and complex project in django and celery backed by postgres, we often came up with testing situations that needed real-world data to work with.
We created an in-house hacked together version of what is now stagDB using only docker and zfs.stagDB allows its user to create a database, load it with data to create a master, and then create as many clones as the underlying hardware may support.
This is still version 0.0.1 of the product. Feedback is welcome.