HyperGrid Thurday took a major step toward realizing its founding vision for a hyper-converged technology adept at facilitating MSPs and enterprise customers spanning out into a multi-cloud world.
New features in the startup's HyperCloud platform include more and deeper integrations with public and private clouds, provisioning of container and big data services on bare metal, and a workload placement recommendation tool.
Those capabilities give customers unified management of applications and infrastructure across HyperGrid's hyper-converged appliance and multiple clouds, with analytics, governance and security embedded throughout, Manoj Nair, HyperGrid's chief product officer, told CRN.
"Our focus and vision is really a software-defined cloud, where capabilities come together," Nair said.
HyperGrid, based in San Jose, Calif., has roots going back further than its official founding in 2016. The company formed through the merger of Gridstore, a hyper-converged appliance vendor, and DCHQ, a developer of cloud orchestration and management software. The acquisition of XOcur last April delivered a business intelligence engine for right-sizing resources.
Nair said the platform was designed not only to deliver a hyper-converged appliance, but to enable the modernization of existing infrastructure with self-service and container orchestration functionality.
The latest HyperCloud release allows decoupling the software from the appliance, making the platform available across a wide range of environments. Several customers had shown a preference for buying their own hardware, or renting public cloud resources, but using HyperCloud as a management automation layer, Nair said.
The release also introduced a stand-alone Software-as-a-Service option for bill analysis and cost optimization.
Roughly half of HyperCloud customers are MSPs using its multi-tenancy capabilities to offer managed cloud services. The comprehensive scope of the platform is enabling VARs to transform their practices to deliver managed services with value-added capabilities, just as larger MSP competitors can do, Nair said.
HyperGrid also added support for Kubernetes, where previously it relied solely on Docker Swarm for container orchestration, and fully integrated the platform with the Puppet configuration manager.
But the most significant new feature is the Intelligent Workload Placement engine.
"We have now the ability for someone already on the public cloud—AWS, Azure, Google—to connect us to your account and we'll start monitoring usage and making recommendations," he said.
The platform will recommend the best environment for hosting existing or new apps as to policy, performance or price. Aiding that effort is a database the company built of performance benchmarks across public cloud instances, giving customers apples-to-apples comparisons, he said.
The latest HyperCloud release adds support of several new public clouds, including Alibaba, IBM's BlueMix, DigitalOcean, and VMware Cloud on AWS. On the private cloud side, the platform now works on environments virtualized with Red Hat's KVM hypervisor and has a deeper integration with OpenStack.
Another innovation was the ability to provision popular big data frameworks like Hadoop, Spark and Cassandra from scratch onto bare metal.
While HyperCloud had previously been able to provision instances across the big three hyper-scale clouds, the latest release adds integration through native APIs for tapping specific cloud services within AWS, Microsoft Azure and Google Cloud Platform.
Michael Boyle, strategic planning director at Westin Building Exchange, a co-location provider in Seattle that offers managed services, told CRN that it sees many customers that are "leery of putting their digital assets on public cloud servers, but do want to leverage a cloud model."
HyperGrid's platform enables those businesses to deploy private or hybrid cloud architectures that deliver "the best of both worlds at a fraction of the cost," Boyle said.
And with the new workload placement capability, Westin Building Exchange can simplify its relationship with customers by proving out the best place to deploy a workload, he said.