While HCI has been a tech buzz-word for at least a decade, the implementation details of hyper-converged infrastructure based systems and clusters are left up to an architect's imagination and can vary widely based on hardware inputs, high availability needs and integrated management components.
At Red Hat, we've been focusing on HCI for two distinct use cases: 1) Red Hat Virtualization with Gluster storage and 2) OpenStack Platform with Ceph storage.
But what if we combined these two HCI use cases and morphed them into a unique hybrid cloud offering consisting of six standalone x86_64 physical systems offering the best-of-breed opensource virtualization, cloud computing, software defined storage, and software defined networking - all with ease of automation, management and performance in mind!?
This talk will cover the architecture considerations, implementation details, and benefits of such a "Hyper Cool Infrastructure".
The hyper-coverged infrstructure (HCI) solution does not have to be limited to just "virtualization infra" or just "cloud infra". When we combine the two and add a few great automation and management components we get the best of both worlds.
A detailed "Hyper Cool Infra" solution architecture will be presented. The minimum hardware, networking, and storage requirements will be discussed.
Automation and management components will be identified with links to automation code.
The pros/cons of the solution will be discussed, especially as it relates to performance managment, component placement and isolation.