What is life like as an SRE in a giant cluster?

mansplainer
João Brito

In the newest episode of Kubicast, we had an amazing chat with Guilherme Oki, an SRE and cloud infrastructure specialist with over a decade of experience. He has worked with fintechs, gaming environments, and is currently working at a stealth-mode startup — and yes, that makes everyone even more curious.
Throughout the episode, we dive into the complexities and lessons learned in large scale environments. Guilherme brings a valuable perspective on what it really means to operate at a large scale, making it clear that it's not just about counting the number of nodes: factors like traffic, latency, and network topology are decisive.
One of the highlights of the conversation is the discussion on geo-distribution and the eternal dilemma of multi-cloud. Is it worth it? At what point is it justified? Using practical examples, he exposes the risks of jumping on the hype train without an actual business need behind it.
We also talked about the pain points and solutions found to scale gaming infrastructure, including historical limitations of EKS and network challenges solved with the adoption of Cilium and eBPF. These tools, increasingly present in the day-to-day of modern clusters, help bring performance, control, and security to highly dynamic environments.
For those interested in cluster Federation, the chat also brings a good dose of realism: not all that glitters is gold, and often the best strategy is to make the cluster more resilient rather than trying to interconnect everything.
The conversation is packed with practical insights and honest opinions on what works (and what causes headaches) when operating infrastructure at a high scale. Whether you are a dev, SRE, or solutions architect: this episode is for anyone who wants to understand the real challenges of the cloud-native world.
Check out the full episode and find out why scaling is much more than just increasing numbers.
Connect with us!
Join our early access program and secure your environment in seconds! https://getup.io/zerocve
🎧 Listen to Kubicast on Spotify as well, and share it with that colleague who will be "unprepared" if the tide goes down :D
Newsletter Getup.
Atualizações sobre Kubernetes e Software Supply Chain Security todos os meses.
Operating Kubernetes in production for more than 13 years. With Quor, this experience extends to software supply chain security as well.
GET UP
© Getup · 2026

