Buying new hyper-converged hardware can seem expensive. CIOs calculate that it is more cost effective to turn to the cloud to host certain workloads. However, this solution is not always suitable. To close the gap between these two options, several providers of hyper-converged infrastructure are now offering consumption-based pricing: the machine is well installed on the premises, but is sold as a service.
In this model, the price for hyper-converged hardware is calculated in the same way as for a cloud platform: A monthly subscription is paid according to the amount of resources actually used. The immediate benefit is that you don’t have to write a large check upfront. We are talking about the Opex model – only the operating costs if and when – unlike Capex – we first buy a good that gets to the capital.
In addition to a simple financial option, this consumption model has other interests. First, the provider usually provides and maintains the machine itself, which simplifies IT and further lowers internal costs. The pay-as-you-go model also significantly reduces the planning effort for CIOs, as they can use the capacity exactly when they need it. Note, however, that suppliers generally require a one-year commitment.
Dell EMC, HPE and NetApp: Three Equal Offerings That Are Not
The difficulty with this model, however, is that each provider has their own idea of how a service is charged. Contrary to what they claim, Dell EMC, HPE and NetApp do not offer any equivalent offerings at all.
At Dell EMC, the program is called Dell Technologies On Demand. It offers three types of consumption: Pay As You Grow, Data Center Utility, and Flex On Demand. Only the latter applies to hyper-converged infrastructures. Companies pay a price per serving node, multiplied by the number of nodes that the builder has installed on site, minus the number of nodes not used in the previous month. Dell EMC installs more and more resources. The company can afford to use a few months more than the theoretical maximum capacity. it then pays a higher price accordingly.
The program has a minimum usage period when companies register for the first time. At the end of this period, they are free to continue using the same hyper-converged infrastructure, return it, or request modernization.
Oddly enough, Dell is vague as to which hyperconverged infrastructure models are compatible with this program. The different areas that Flex On Demand covers are likely to vary from period to period.
At HPE, the consumer billing program is called GreenLake. But it’s much more than just paying. Rather, it is about subscribing to the support of the internal consulting company PointNext. In return, PointNext offers to implement the solution and keep it up and running. And with “solution”, PointNext means “total solution” with software and services from third-party providers that normally go hand in hand with hyper-converged simplicity infrastructures. Note that GreenLake was recently referred to as GreenLake Flex Capacity and that one or the other name will prevail in the catalog depending on the application. It doesn’t matter: service and prices are the same.
In fact, the subscription billing depends on the capacity of the infrastructure deployed on site, as well as payment options that vary based on the measured amount of actual resources used. Unlike Dell EMC, HPE can update the existing infrastructure or add additional nodes at any time. In addition, it is possible to enter into a multi-year commitment with HPE.
At NetApp, the consumer accounting program is called Cloud Consumption for NetApp HCI. It is officially designed to provide businesses with a monthly billed private cloud deployment solution. In fact, NetApp has embraced the concept of the cloud to such an extent that its HCI platform no longer means HyperConverged Infrastructure, but Hybrid Cloud Infrastructure.
When a company signs up for the Cloud Consumption program, NetApp installs a cluster of HCI H400 or H600 nodes in its data center and charges based on the number of nodes. This cluster must contain at least four storage nodes and two compute nodes. The obligation is one year. Beyond this period, it is possible to pay from month to month and expand the cluster.
The responsibility for managing this or that component in the solution is not very clear. When Cloud Volume Services are deployed with the HCI cluster, they are managed by NetApp, but the company must take care of the underlying infrastructure. As in the case of Dell EMC, these contract details seem to change from time to time.