Free DownloadWhat is hyperconverged infrastructure? Guide to HCI
Hyperconverged infrastructure is a software-centric architecture that brings compute, storage and virtualization resources into a single, highly integrated system. Let's take a closer look at hyperconverged infrastructure technology, including use cases and implementation, tradeoffs, current vendors and product offerings and a look ahead to the future of the technology.
Hyper-converged cloud key to private cloud success
Granular scaling and rapid deployment help make hyper-converged infrastructure the ideal platform for building out, deploying and scaling out a private cloud.
One promise of the cloud is a fast time to value, and another is rapid scalability. Self-service provisioning by IT consumers usually takes minutes, leading to a near-instant response to business needs. While scaling cloud workloads fast means immediate responses to changes in a workload, such as is required with a new product launch, both of these advantages are hard to achieve with a private cloud, as there's seldom a lot of predeployed capacity waiting for new workloads.
No public cloud is built overnight, nor is a private cloud implemented in a single day. Underneath any cloud infrastructure, public or private, is a virtualization platform that delivers compute resources.
In large public cloud data centers, capacity is deployed as multiple racks of servers that are, in turn, deployed and configured using in-house automation. A private cloud seldom has the scale to develop the entire automation stack, and it often relies on hardware vendors to provide automation tools.
HCI products are designed to be deployed fast, taking hours rather than weeks to build a full virtualization infrastructure. They also allow for a small start; as few as three physical servers can be deployed at the start of hyper-converged buildout.
Each HCI node includes storage and compute, reducing the need to involve additional teams and expertise to deploy the infrastructure. By speeding up infrastructure deployment, a hyper-converged cloud platform can enable faster implementation of a private cloud. Faster implementation enhances the ability of IT to prevent the use of the public cloud for shadow IT implementations.
Software-defined infrastructure
Cloud management platforms typically require integration with provisioning tools for the compute and storage components of a virtualization platform. Developing custom integration can take weeks if the virtualization infrastructure isn't designed for software control, and automating GUIs and command lines is slow and error-prone.
In addition to faster deployment, HCI platforms, as software-defined infrastructures, are designed for software control. Most provide a REST API that exposes all the management features, making integration faster and more consistent. This software-driven design behind the hyper-converged cloud simplifies integration between a cloud management platform and the virtualization platform, further shortening the time required to deploy a private cloud infrastructure.
Scale-out expandability
One objection to the idea of a private cloud is that it's impossible to have the elasticity of a public cloud. Elasticity is the ability of tenants to expand and contract their consumption of cloud resources as their workloads change. Expansion is limited by the free capacity of the underlying virtualization platform.
A hyper-converged cloud infrastructure can accelerate virtualization deployment [and] shorten private cloud adoption time.
Neither public cloud operators nor enterprises want to have a lot of spare, unused capacity, as that's money spent not generating value. Public cloud platforms can pool the remaining capacity across multiple tenants and reduce the total amount of unused capacity; private cloud platforms don't have this luxury.
Private clouds do benefit from being more aware of business changes that drive the need for elasticity, though, and can therefore be mindful of demand changes earlier than public cloud providers. In a private cloud environment, large-scale elasticity is provided by acquiring and deploying new physical hardware.
HCI platforms allow for expansion by as little as a single node, a routine activity that may occur multiple times per year or month. Furthermore, expanding an HCI cluster adds compute and storage at the same time, which can help to avoid bottlenecks as the infrastructure scales.
The software interfaces to the hyper-converged cloud infrastructure platform to abstract away the physical nodes, so there's no change in management when scaling from an initial three nodes to a cluster of a dozen HCI nodes. This eases private cloud elasticity by enabling pure scale-out growth.
The hyper-converged cloud
Business units today are unlikely to be willing to wait months for conventional virtualization platform deployments. A hyper-converged cloud infrastructure can accelerate virtualization deployment, shorten private cloud adoption time and -- as private cloud adoption grows -- can scale out by adding additional nodes without changing cloud platform management. Rapid deployment and the ability to granularly scale combine to make HCI the ideal platform when building out a private cloud.
