August 3, 2018
The most common use of x86 Servers today is to run hypervisor software that enables the hardware to be shared by multiple virtual machine (VM) instances.
This type of architecture has dramatically changed the economics for every business in a positive way for close to two decades. However, there have been some challenges when trying to scale beyond a single hypervisor.
The Evolution of Hypervisors
Some quick history here – since the beginning, the greatest challenge for hypervisors was the ability to scale beyond a single host due to storage sharing limitations. Although the central processing unit (CPU) and memory resources can be leveraged across multiple hypervisors’ hosts to run dozens and hundreds of VMs, the local storage cannot be natively shared even up to this day.
To overcome this local storage issue, system administrators started using SAN and NAS solutions early on. This worked out well and many data centers began to deploy 2-node or 3-node hypervisor clusters.
Everyone was happy, until the SAN costs began to rise exponentially.
Read the entire article here, How to Convert Your Hypervisor Into a Fast HCI Appliance
Via the fine folks at DataCore Software.