We were recently interviewed by Chris Kanaracus for the leading tech publication TechBeacon for their article How they did it: 6 companies that scaled DevOps. Here is an expanded version of the interview, covering the state of DevOps in 2018:
Q: In 2018, what are the common themes that mark enterprise-wide DevOps deployments? Successful ones, that is. :)
We're seeing a spectrum of approaches to DevOps in large enterprises. Some places are stuck with large "DevOps" teams busy with all kinds of automation but not really facilitating joined up communications between Dev and Ops - the Anti-Type B in the DevOps Team Topologies patterns (https://web.devopstopologies.com/). These organisations have decided that "DevOps" is mostly about automation, and so are moving forward but not as effectively as they could and should be. Other organisations are adopting the SRE pattern from Google (Type 7), again with mixed success. For SRE to work well, the SRE team needs to be able to say "no" to low quality code coming their way, but some organisations need to do more work to build more trust to make that work. The SRE pattern seems to be gaining traction in heavily-outsourced situations with multiple suppliers, so I expect we will see some new "SRE as a Service" offerings emerging over the coming years to meet the need for highly-skills operations & reliability engineers in the enterprise.
Q: How would you grade the maturity of DevOps as a discipline today? And the same for its awareness level?
There is a broad range of maturity and awareness around DevOps. Some people think that DevOps is just infrastructure automation and deployment pipelines, which is such a lacklustre view. In some organisations it can require a lot of to get proper ongoing collaboration between "teams that build" and "teams that operate". However, we're also seeing significant success stories from large organisations across many different sectors, so there is really now no excuse for not adopting tried and tested team collaboration and organisation design patterns to make DevOps work well.
Q: What's lacking in DevOps practices/culture today? If you were assigned to be the grand architect for DevOps best practices, what sort of things would you implement?
In many organisations there is still a lack of focus on core engineering practices (proper Continuous Integration, inter-team coordination, API design, test data management, etc.) - things that are addressed by Continuous Delivery practices (in the 2010 book by Jez Humble and Dave Farley). I also see a significant lack of awareness of software operability techniques and practices (using logging and time series metrics effectively, collaboration on operational aspects between teams, testing for operability in a deployment pipelines, etc.). I find that a combined focus on Continuous Delivery and operability practices produces good results.
Q: What factors cause enterprise-wide DevOps efforts to fail, or at least flounder?
Lack of leadership buy-in will usually prevent DevOps initiatives - the leadership must want to improve, to go faster (safely), to change the culture within the organisation. However, DevOps transformations can flounder if we focus too much on the technology and not enough on flow, feedback, and metrics. There is a danger for some companies of hiring "DevOps" engineers who then go on to spend months building some complicated tooling thing which does not align to the value delivery stream.
Read the full article on TechBeacon How they did it: 6 companies that scaled DevOps