DevOps and the Database Bottleneck

Databases are central to just about every application system but remain a neglected topic when it comes to DevOps. Seriously - while application change delivery has evolved to where changes can be applied in amazingly short intervals, most enterprise databases still live in a silo that exists somewhere on the other side of a bureaucratic ticket-driven process from 2003. To be fair, databases are critical things which, when broken, will break everything else. However, given that you can never be faster than your slowest operation, this central part of application systems is often a drag on teams’ overall throughput. So, something has to give if we are to get through the Database Bottleneck bring agility to application delivery and the businesses it supports.

The good news is that there is an emerging body of knowledge on how to conquer this challenge and this talk will:

  • Describe the problem of automating / accelerating database changes
  • Explain how the problem impacts the various roles in app delivery
  • Introduce the two main approaches to database change automation - state-based and migration-based
  • Explain the strengths and weaknesses of each approach
  • Discuss how they fit into CI/CD workflows

This talk will discuss problems, approaches, and solution patterns It will explicitly not have a tool or demo focus, but instead look at the approaches, processes and context into which you can fit tools.



Dan Zentgraf


Dan is a technology professional focused on bringing DevOps practices to enterprises. He has over 20 years in the IT space and has served in a variety of capacities including Product Management,