Kubernetes at the Edge - How Roche is on the way to manage 1000+ clusters
summary
Roche runs Kubernetes on-prem in labs to avoid cloud privacy and latency limits and aims to scale toward 1000+ clusters by treating Kubernetes as the OS layer. The stack uses Talos OS (immutable, no SSH), Cilium networking, GitOps for repo-per-cluster scaling, declarative in-place updates and multi-node HA, plus secure boot with TPM and full-disk encryption for physical-device protection. Deployments currently host mostly stateless workloads, remote debugging and backups remain open operational problems.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:01Talk agenda
- 2:06Ro divisions and healthcare
- 5:18Edge computing for Ro
- 8:47Software deployment challenges
- 10:32Why edge computing is hard
- 13:26Kubernetes at the edge
- 16:59Platform complexity and automation
- 17:21Talos OS for edge Kubernetes
- 18:42Secure boot and disk encryption
- 21:05Closing remarks
- 21:26Question and answer session
related talks
30mEdge Kubernetes and Healthcare: A Next-Gen Architecture that Scales Across Thousands of Locations
10mKubernetes: Enabling Edge Workloads
17mKeynote: Extending Kubernetes: Application Orchestration to 1000’s of Clusters in...
33mThis won't hurt a bit: Taking Kubernetes to thousands of dental offices
12mHow to Effectively Manage Kubernetes in a Regulated Environment
12mRook: Why Would You Ever Deploy Ceph Inside Kubernetes?
36mKubernetes in the Enterprise Transforming Rabobanks On Premise Applications
10mDevelop once, deploy anywhere: flexible infrastructure extends to the edge