DEV Community

Page It to the Limit

Shifting to Full-Service Ownership With Leandro Pinto

Implementing a Full-Service Ownership Model

Leandro and George discuss a few of the practical considerations that teams need to tackle when moving to a model where teams own the code they ship to production.

Building Ownership Teams

We discuss the transition from just writing software code to empowering teams to own more of that lifecycle. Leandro tells us a bit about how his company, MessageBird, has approached this problem.

“It’s giving ownership to both functional and non-functional aspects of an application to a single team: functional aspects being something like delivering product features to a customer, and non-functional being aspects like availability, performance, and security.”

But developers don’t want to be on-call

Leandro debunks two popular myths about going on-call.

“You hear a lot of people saying that developers don’t want to be on-call. You’d be surprised how many people step up to and take that responsibility when you actually empower them to fully own their services. People will take that responsibility and go the extra mile.”

Practical ways of removing roadblocks

As teams shift to a full-service ownership model, one of the biggest challenges is shifting roles and responsibilities in a way where things don’t fall through the cracks. How exactly is that done when people have a finite amount of bandwidth and knowledge? Do you hire more people? Do priorities just change?

“We still have a platform team that is more focused on [infrastructure]. But we also invested a lot in training. So instead of having a skills gap, our teams have an intersection with that [platform] knowledge.

Unexpected lessons learned

We talk about both the challenges and the things that became much easier when MessageBird transitioned to a full-service ownership model.

“What makes it easier is the closeness you have to [customers] and how easy it is to get to the bottom of issues with your service. It brings people closer to the punchline of solving problems.”

Incident alerting

We discuss how this shift in thinking is being driven and reflected by how MessageBird manages their approach to incident alerting.

“In the past we structured ‘services’ around escalation policies and team structures… Now our services are at the center. Now our on-call rotations and escalations are set up around who actually owns that service, instead of reflecting our organizational structures.”

Ops in my Dev or Dev in my Ops

We talk about the fear most teams have that software engineers can’t own running code in production. The first wave of DevOps seemed more focused on having Ops engineers learn to code. The second wave we’re in now, is about Devs learning how to Ops.

“What I always say is, honestly, software engineers have a great foundation for what they need in order to be successful when running software in production.”

Recurring Questions

George and Leandro both get real about the mistakes we made early in our careers.

References

Episode source