This has been bugging me for a decade now.
Architectures such as Kubernetes allow for the orchestration of heterogenous hardware to run many types of applications using HTT-Protocol. But there's no common architecture that allows for the orchestration of heterogenous hardware to run Linux desktop applications using the Linux kernel's local machine management protocol. Is there?
Let's say I want to spin up six old laptops, on a W/LAN, and have my local kernel abstract over all their CPUs as if they were running on the same machine. (Nevermind for a second, how slow this would be, due to the memory siloes.)
What's the shortest path to coding this up?
No comments :
Post a Comment