What is Hyper-Converged Infrastructure?
Hyper-Converged Infrastructure
First, there was storage management that relied on fixed lists of service requirements (which, once implemented, could not be changed), proprietary hardware, management software from multiple vendors, and specialist staff. It was expensive, wasteful, and overly-complex.
Shared storage that behaves just as storage provided by SAN or NAS devices does, is created by the ESXi server which pools together direct-attached storage devices (not external storage systems, as with storage virtualization) from across the cluster.
Hyper-converged infrastructure can be achieved in one of two ways. Most commonly, third‐party storage software is run in virtual machines that sit on top of a hypervisor. While this method does achieve hyper-convergence, it results in the use of too many resources, as well as reduced performance and less than optimal integration with the current environment.
The second, more innovative, method – which is the method used by VMware – is to build the storage software into the hypervisor itself. Whereas in the first method, convergence takes place on the hypervisor, in this second method, convergence takes place inside the hypervisor.