Every time a product owner says “We should pull in XYZ data to the mobile app” or “we need a new app to address this healthcare fail” an engineer needs to make two crucial decisions:
- Where and how should the new code acquire data the company currently has?
- Where and how should the new code record the data that will be newly created?
Sadly, the most expedient answer to both questions is “just use one of our existing databases”. The temptation to do so is high when an engineer need only add a database migration or write a query for a familiar database. The alternative might involve working with the organization’s infrastructure team to plan out changes to the operational footprint, then potentially making updates to the developer laptop setup.
Decisions expedient for today aren’t necessarily the best decisions for Rally’s long term delivery velocity. We recognized database reuse and sharing was fairly common at Rally, so we tried to stop to the practice in Spring 2017. We were concerned the company’s development speed and agility would eventually grind to a halt.
Read more at Rally Engineering