BYO Infrastructure: Who needs to run a control plane, anyway?
Thank you to Mirantis for sponsoring the Open Infrastructure Summit!
OpenStack has constantly evolved to catch up with VMware’s 10 YEAR head start. Containers solved lots of problems, but not all. When Kubernetes replaced the use of containers alone, real advancements in scaling and self healing came to be.
But now that we have a fully orchestrated, auto-scalable, self-healing control plane that can run anywhere on any infrastructure, why do we need to have the control plane and the rest of the physical service coresident? Why don’t we let someone else manage the updates and upgrades in a public cloud, while WE focus on the DATA portion of the services (compute and storage) on prem? We can use a service mesh (which WE control) overlaying a VPC network to apply all of our corporate rules and constraints, while the headache of deploying, managing, upgrading and maintaining the infrastructure can be left to experts in such things.Follow the transition from baremetal to virtualization to containerization and beyond.
linux foundation
