The HP ConvergedSystem 200-HC EVO: RAIL announcement indicates that VMware EVO: RAIL is now a mainstream option in the hyper-convergent market. Keith Townsend explains why.
HP made a splash during VMworld 2014 Europe by announcing an EVO: RAIL product called HP ConvergedSystem 200-HC EVO: RAIL. EVO: RAIL is VMware's hyper-converged OEM program; it lets OEMs partner with VMware to provide an easy to deploy 4-node vSphere cluster that includes all the storage and compute in a 2-U form factor. These systems compete against solutions from the current hyper-converged mindshare leader Nutanix.
Common wisdom dictates that HP would have wanted to create a hyper-converged system using all HP technologies. HP has all of the individual parts needed, including StoreVirtual (Lefthand acquisition), ProLiant server hardware, and a long list of orchestration and infrastructure management software solutions. I've argued that HP and Cisco were resisting the EVO: RAIL program because of the desire to promote their competing data center management visions.
Cisco's data center management vision has UCS Director at the center. With tight integration with ACI, Cisco's data center management strategy is compelling; likewise, HP has been preaching the virtues of an HP converged data center. Both Cisco and HP's visions for the data center cater to large enterprises. HP advertises the ability to set up a converged infrastructure in less than 20 days. A 20-day deployment is compared to hyper-converged systems that measure installation and configuration times in hours vs. days.
HP's existing Converged infrastructure is positioned to support large-scale deployments and compete with the VCE Vblock. Legacy HP Converged and Vblock include full-featured block storage solutions and networking; the tightly integrated solutions are designed to scale to support thousands of virtual workloads.
In contrast, EVO: RAIL solutions scale up to approximately 400 workloads; each appliance supports 100 workloads. HP ConvergedSystem 200-HC EVO: RAIL can scale up to four appliances for a total of 16 nodes. Enterprises looking to scale beyond 400 workloads would be looking at one of HP's larger converged systems or a VMware EVO: RACK. (EVO: RACK is still in development, and it's unclear if HP will offer an EVO: RACK version of its converged system.)
Now that VMware has two of the largest server vendors in the EVO: RAIL camp, it will be interesting to see if this places pressure on Cisco to offer an EVO: RAIL solution. Even without Cisco, VMware EVO: RAIL has an impressive list of OEMs, each of which offers a different level of value over the base solution. EMC, for example, includes RecoverPoint replication as part of the EVO: RAIL solution.
If you're considering a hyper-converged solution, with HP onboard, it's safe to say EVO: RAIL is a mainstream option.