How Hybrid Cloud Ended the Infrastructure Debate for Good

Why Hybrid Cloud Makes Sense for Many Businesses

Welcome to the world of hybrid infrastructure. As they grow, businesses that once managed data on-premises find it makes more sense to expand into the public cloud than to continue to buy and operate more hardware. The scalable and resilient nature of the public cloud, combined with its easier management and access to technology such as artificial intelligence (AI) platforms, delivers the agility businesses crave.

At the same time, on-premises data centers and private clouds have their place: They can provide better performance, reduce latency and support data privacy and other compliance requirements.

“There comes a time when they are an intersection where a blend is best,” says analyst Matt Kimball of Moor Insights & Strategy.

This inflection point depends on the unique needs of each business, and budget can play a role. For example, a company that starts on-premises but wants to take advantage of AI may find it easier and more cost-effective to train their AI models in the cloud, where they can rent expensive graphics processing units (GPUs) instead of buying them, Kimball says.

In contrast, startups may begin in the cloud because they can’t afford their own infrastructure. But as they mature, it may be cheaper to run some workloads on-premises, he says.

RELATED: Take these three steps to maximize your hybrid cloud environment.

How to Simplify a Hybrid Cloud Migration

4Wall’s migration to hybrid cloud was relatively easy because it had previously deployed Nutanix’s hyperconverged infrastructure equipment and software. HCI combines servers, storage, networking and virtualization into a small-footprint appliance and is managed through a centralized software tool.

Last year, 4Wall upgraded its Nutanix software licenses to Nutanix Cloud Clusters, which allow it to easily migrate virtual machines and data to Azure without having to refactor or rearchitect applications, Adams says. 4Wall’s IT staff can manage the on-premises and cloud environments through a single management console.

“It took us less than two weeks to integrate to the cloud,” he says. “We can now lift and shift applications and move them back and forth from our data center to the cloud. It’s seamless.”

4Wall, which has seen its employee head count quadruple to 1,000 in the past four years, is still fine-tuning its application strategy. In recent years, it standardized on Software as a Service (SaaS) apps, including Microsoft 365 and Asana, but it still has between 100 and 200 virtual machines.

The IT staff plans to move static VMs to Azure. The plan is to grow further with company acquisitions; as the company does, Adams will likely shift some other applications from its core data center to Azure. “I don’t want to buy more servers. Now, I can just scale up in the cloud,” he says.

How EchoStar’s Cloud Journey Delivered Agility

EchoStar historically relies on in-house data centers, but a few years ago, the satellite communications company embarked on a cloud journey, moving apps and data to Google Cloud and taking advantage of its machine learning (ML) capabilities.

This effort has led to more innovation, better services and more efficient and cost-effective operations, says Mridul Malayanil, a scientist and senior director at EchoStar.

To migrate workloads to the cloud, the company deployed Google Anthos, which allows enterprises to run applications in their own data centers or the public cloud using Google’s Kubernetes containerized software, he says.

Malayanil’s team runs business service systems, which includes software for managing customers, partners, sales agents, installers and the dealer network.

The Englewood, Colo.-based company refactored many apps to run on the Google Kubernetes Engine Enterprise environment, and today, nearly all of EchoStar’s applications and all of its data are in Google Cloud, he says. The company has also migrated VMs to a private cloud within the Google Cloud VMware Engine.

How Cloud Makes AI Easy to Access

EchoStar’s first cloud use case was with ML applications. The company initially bought GPUs to try ML in-house but found Google Cloud more cost-effective. Now, it uses ML to do more advanced analytics, better understanding customer preferences and personalizing sales offerings based on demographics such as income and location, Malayanil says. 

EchoStar is also testing a generative AI chatbot with internal users to help with customer troubleshooting. They can ask questions and get quick answers through EchoStar’s knowledge database, he says.

The company is consolidating data centers across the world and eventually hopes to migrate everything to the cloud. However, today there continue to be dependencies and technical constraints that require the company to keep its hardware.

Employees still need to produce hundreds of business and financial reports using existing on-premises tools, Malayanil says. Users can’t access data directly from Google Cloud. So, the company replicates the data from the cloud to on-premises data marts in real-time. 

“Agility and the ability to scale in the cloud is big, and our ability to respond to business needs has been much faster,” he says. “And cost-wise, we’ve saved a lot of money.”

How to Increase Business Resilience with Hybrid Cloud

Strike, a Spring, Texas-based company that builds pipelines and facilities for the energy industry, is modernizing IT infrastructure this summer with new Dell Azure Stack HCI equipment.

The solution, which features Microsoft’s Hyper-V virtualization software, software-defined storage and networking and Azure Virtual Desktop, allows the company to run virtualized Windows workloads in-house while connecting to Microsoft Azure cloud services. 

In doing so, the 600-employee company will improve resiliency thanks to the Azure cloud, says Strike IT Infrastructure Manager Greg Altman.

The company purchased the new HCI equipment because existing hardware had reached its end of life and because it wanted to move away from its existing virtualization software vendor.

Strike is consolidating two data centers into one. The new Dell Azure Stack HCI equipment has five nodes, so if two nodes go down, it can still run essential operations, Altman says.

The company will store backups of apps and data in Azure. If its data center gets knocked out, Altman can quickly turn to the Azure cloud to keep its enterprise resource planning (ERP) software, SQL Server databases, timekeeping software and other essential applications running. The IT staff will use Azure Site Recovery to spin up key applications in an emergency.

“We will have enough redundancy that we’re not going to have a single point of failure,” he says.

Strike, which began implementation in May, expects to complete the project in August. The company will use Azure Arc, a central management tool for managing its on-premises and cloud environment.

Besides Azure, the company has replaced some on-premises apps with SaaS tools, such as Microsoft 365. The company will also use some Azure services that will integrate different SaaS solutions to its on-premises ERP software improve business workflows, Altman says.

“We’re moving services to the cloud where it makes sense,” Altman says. “If there is an Azure service that will accomplish a goal, and it’s cheaper than building our own solution or finding a third-party app, then we will fire up the Azure app.”