By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
VMware officials also gave some details on the likely scale of the larger EVO:Rack product and suggested the EVO product line would offer scale-out capabilities.
The addition of HP – the world’s largest supplier of servers – and HDS to a list that includes Dell, EMC, Fujitsu, Inspur, NetOne and SuperMicro means VMware’s EVO:Rail can now point to partnerships with the makers of the bulk of servers supplied worldwide.
EVO:Rail is a hyper-converged appliance that marries compute and storage in an appliance optimised for VMware environments.
Hardware compute components are strictly specified by VMware to its server maker partners.
EVO:Rail runs VMware software, namely vSphere, to virtualise CPUs; VSAN storage virtualisation; vCenter Log Insight; and the EVO Engine that handles deployment, configuration and management of resources. The software components are knitted together by application programming interface (API) calls.
The product is formed of base unit nodes that comprise a maximum of three 1.2TB HDDs, plus 400GB of flash as a cache. Four of these nodes make one appliance and four appliances can be linked to provide an 8U cluster with just over 50TB of storage.
More on VMware EVO
That type of capacity puts EVO:Rail in the small to mid-sized customer bracket, with each appliance catering for 100 virtual servers or 250 virtual desktops.
EVO apparently competes with existing converged stack products, such as the EMC/Cisco vBlocks, and FlexPods, the NetApp equivalent. But those products also contain networking components, so will that be added to EVO?
“We’re very sympathetic that networking and virtualisation are very important to our customers,” he added. Look out for the addition of VMware NSX network virtualisation to the EVO stack at some point.
McDonough also gave some details about the proposed big brother to Rail, EVO:Rack. This would comprise full racks of EVO architecture to provide 128TB or more.
Asked whether it is planned that EVO would be a scale-out architecture where the addition of nodes provides seamless expansion of compute and capacity, McDonough said that currently each cluster is a discrete management domain with communication between them via a top-of-rack switch. But when asked about scale-out capability, he said there would be “great news in future”.