Lurid tales of failure, heartwarming stories of human triumph and big visions on the future of DevOps, whoever dares claim that going cloud-native is dry has yet to see our top picks for this year’s KubeCon Europe in Barcelona.

As always, there are lots of cool talks by industry leaders and tons of content. So, to help you weed through the names and hype, here’s our list of top 4 talks.

And in true clickbait fashion, number 4 is the one not to miss!

Kubernetes Failure Stories and How to Crash Your Clusters

Who doesn’t like a juicy story of failure? It satisfies our need for schadenfreude, but crucially, is a great opportunity to learn. This session by Henning Jacobs from Zalando offers an introspective, humorous look at how to avoid the pitfalls that Zalando has already tread in, both technical and cultural, while adopting Kubernetes across 200 engineering teams. In all, attend the session to bend your unknown unknowns into known unknowns!

Grow with Less Pains - Meshing from Monolith to Microservices

We’re all in the game of moving away from monoliths in some way, shape, size or form. OfferUp is no exception. Leo Liang and his team used L5D, Consul, Nginx and Prometheus to build up their own service mesh in their world of EC2 instances, ECS, Kubernetes and Lamdba. They’ve been very open and outspoken about this subject in the past and have posted a handy three-part blog series on their growing pains.

Reliability Engineering for Humans

Humans. Don’t they always mess up your grand plan for transitioning to microservices, DevOps and an amazing engineering culture? In this session, Hannah Foxwell will discuss a more personal approach to release engineering through a human lens. She combines site reliability engineering with HumanOps and explains how to use SRE practices to improve the health and well-being of your team.

A Tale of Two Worlds: Canary-Testing for Both Machine-Learning Models and Microservices

Once you’ve figured out a way to release microservices frequently without crashing a whole dependency ecosystem, you can start asking: what happens when you add data models to the go-to-production mix? Can data science learn from the microservices space? In this session, Mesosphere’s Jörg Schad and Vamp’s very own Vincent Lesierse show how experiences learned from releasing microservices on Kubernetes can be applied to deploying machine-learning models without impacting users.


Our top 4 covers some of the stepping stones on any cloud-native journey, from just figuring out migration, to having a smooth CD release pipeline in place. Our next blog installment will unpack how to handle smooth release through canary-testing, so subscribe or stay tuned. At Vamp, we eat, sleep and breath containerization to cloud, so it’s our mission to share knowledge wherever we can.