Awesome agile development teams have known for years what TDD and collaboration can do for their performance and their quality of life. The DevOps movement and its tooling can bring that to infrastructure teams as well.
I have observed an interesting pattern on several recent software teams that strove to adopt infrastructure automation: after a brief honeymoon period, ~90% of the developers came to dislike much of the tooling around infrastructure deployment and the activities around managing it. They wanted to offload that work to a different team so that they wouldn’t have to deal with it at all—in effect, requesting a new silo to separate themselves from a class of work and responsibility(!).
At first I was surprised. Why this distaste for tools and practices that actually gave them greater feedback on and influence over the runtime quality of their apps? It turns out—unsurprisingly—that there are some serious differences between the state of the art in app-level tooling and the much less mature world of infrastructure tooling.
As I recognized the pattern I also wondered: how do we overcome this, so that teams can take their next steps towards that greater achievement? This session aims to help.
In this interactive talk, I will share my own experiences, observations, and insights as well as inviting audience members to share theirs. Attendees should walk away with a stronger understanding of some of the particular challenges of developing infrastructure, and with an assortment on tips for how to deal with those challenges—both in tooling and in technique.