Bypassing the network stack with eBPF
Recent research has shown that Kubernetes networking can be sped up significantly with a small effort. How hard could it be to recreate those results?
I’m currently working on infrastructure at OpenAI and writing about coffee and code.
Recent research has shown that Kubernetes networking can be sped up significantly with a small effort. How hard could it be to recreate those results?
Recent research has shown that Kubernetes networking can be sped up significantly with a small effort. How hard could it be to recreate those results?
Follow my journey (and learn from my mistakes) in redesigning and developing my new portfolio / blog using Astro, Tailwind, and as little JS as possible.