What Will Be Hot at KubeCon? Platform Engineering, of Course
What will be hot at KubeCon this year?
Here are three topics that the Cloud Native Computing Foundation’s Taylor Dolezal says are hot, hot, hot!
GenAI/LLMs are hot! This year, KubeCon has an unconference style, AIHub.
Generative AI and LLMs are generating a lot of interest, especially when considering the security of large language models, Taylor Dolezal, Head of Ecosystem at Cloud Native Computing (CNCF) said. Then, there is all the legacy infrastructure that companies have. How can old hardware be leveraged for LLMs? And how do LLMs fit with Kubernetes?
Platform engineering is hot! There are more than 25 sessions about platform engineering at KubeCon Chicago. Holy cow!
But what is it? The definition of platform engineering is starting to congeal. People are starting to get a better idea of how site reliability engineering and other topics fit into what we think of platform engineering overall.
What’s interesting for Dolezal? Platform engineering is becoming a way to create a space for internal product teams.
Essentially, rather than just being a platform team that runs a specific platform and bottlenecks, the organization, focuses on how to make a good product that you can proliferate throughout your organization,” Dolezal said. “In previous roles that I’ve had, that’s the approach that we’ve seen be really successful. So really, I’m enthused to hear some folks and their journeys on that front versus being, you know, running a global logging cluster, you know centrally located, or just within one cloud, or a few others. Hearing these SRE principles poured into a platform, and some of that product management, I think, is really interesting.”
WebAssembly is hot! KubeCon has a full day dedicated to it as part of a co-located event on Monday, Nov. 6.
WebAssembly is growing up, Dolezal said. Most uses are on the frontend. It is getting close, though, to that “containerization moment,” that Docker had. But what’s cool about WebAssembly? It can complement containers, especially in edge computing use cases.
“Or if you want to parse something out or keep your data contained within a specific network or area within your organization, you can preprocess that and work with that before it actually makes it to the cloud, which is really useful as well,” Dolezal said about WebAssembly. “So you can start to see these architectures really evolve and become even more multilayered than they are.”
