Don't blame it on Kubernetes! | Federico Nardini
From beginning of January 2020 to end of March 2021, I lead a Reliability Engineering team of the UK Government Digital Service. Our responsibility was to run and maintain a platform based on Kubernetes architecture. What we actually ended up doing over those 15 months, was to retire the platform. Let's look together at our journey: incidents, recoveries and more or less planned migrations, affected by the loss of exceptional people on the way. Key audience takeaways: - Why inadequate top down decisions and optimistic commitments are "expensive" - How I came to the realisation that the Kubernetes platform was set for failure in first place - Why Kubernetes as a technology was not the real problem here - What did we end up doing and how did we serve our users