When more than 22,000 soccer fans fill PNC Stadium to watch the Houston Dynamo FC, service outages are unacceptable. That’s why the club runs most of its workloads in an on-premises Nutanix private cloud environment. Photography by Bob Stefko.
When It Comes to the Cloud, Each Company Must Choose Its Own Path
Businesses blaze their own trails, in and out of the cloud, to meet their unique needs for data management.
The benefits of the cloud are so well known by this point that most IT leaders can likely rattle them off in their sleep: scalability, agility, availability and lower upfront costs, among others. So when the Houston Dynamo FC, a Major League Soccer team, moved its workloads from the public cloud back in-house to a Nutanix private cloud environment (with its disaster recovery in the Nutanix Xi Leap private cloud), it’s fair to say it was following a path less traveled.
But the Dynamo’s IT leadership is delighted with its decision, noting that going forward, most of its workloads will be on-premises.
“One of the big things for us is that we have pretty strict response time requirements,” says Frank Arnold, the team’s vice president of IT. “When we’re running an event, any sort of an outage that we’re not in control of can be very detrimental to the services that we run. That was a big piece of pulling applications back on-premises — so that even in a worst-case scenario, if we were cut off from the internet, we would still have that availability.”
Chris Kanaracus, a research director in IDC’s infrastructure practice, says the Dynamo’s decisions shouldn’t come as a surprise. Indeed, a hybrid approach to the cloud is “here to stay,” he says, with 97 percent of organizations already using some mix of on-premises and public cloud resources.
However, while most IT and business leaders know that the cloud can work for them, many — including those already using cloud resources — still aren’t completely sure about the optimal approach for their own organizations. Kanaracus cites several challenges facing IT and business leaders as they figure out the specific mix that will work best for them; for example, applications must be optimized for the environment where they run, cloud management can become complex, and organizations risk overbuying resources.
“Hybrid cloud is a double-edged sword,” Kanaracus says. “It increases flexibility but adds complexity, especially when you’re talking about an enterprise’s procurement and cost management practices. There are a lot of vendors out there that are selling tools to help enterprises manage this problem. But as a customer, you’ve got to closely evaluate which ones will work best for you.”
If we suffer a fiber cut to the building, we still need to be able to sell and scan tickets.”
Frank Arnold Vice President of IT, Houston Dynamo
For Some Workloads, Downtime Is Unacceptable
The Dynamo run the same sort of back-office applications that most businesses run. If those went down for the bulk of a work day, it would be inconvenient. But the organization also has game day technologies — such as a ticket scanning application — that the club simply can’t afford to lose. “If we suffer a fiber cut to the building, we still need to be able to sell and scan tickets,” Arnold notes.
In fact, the team has suffered fiber cuts in the past due to nearby construction, although those outages never affected game day operations.
The team is investing in hyperconverged infrastructure from Nutanix, placing nodes at three locations: the team’s stadium, a nearby corporate office and a training facility. For some operations, such as those that rely heavily on video, the organization is running workloads as close as possible to endpoints — another advantage offered by the on-premises infrastructure.
Businesses should place workloads where they make the most sense, rather than building an environment to suit a certain preference, Arnold says. “You have to consider what is important to your organization.”
RELATED: Learn how innovative tech can help drive business outcomes.
How to Increase Efficiency by Eliminating Unnecessary Hardware
When Rose Gilligan became senior director of IT for the American Association of Nurse Anesthesiology in 2019, she knew that something had to change about how the organization was managing its data. Over the years, the AANA had accumulated so many physical and virtual servers (150) that they outnumbered the employees (130).
“It was overwhelming,” Gilligan says. “We hadn’t cleaned anything up, and we were starting to reach the point where we needed to replace the assets, not to mention consolidate.”
Gilligan determined that AANA’s best course of action was to migrate nearly its entire environment to the Amazon Web Services public cloud, getting itself out of the data center business entirely. CDW led the migration effort, in some cases simply replicating on-premises resources in the public cloud through a lift-and-shift, but often replatforming infrastructure to avoid over-provisioning resources.
Rose Gilligan moved the American Association of Nurse Anesthesiology onto Amazon Web Services’ cloud platform after becoming IT Director in 2019, taking the organization out of the data center business entirely.
“Where there were multiple virtual machines for basically the same server or the same workload, we combined them in a way that reduced costs for both storage and licensing,” says Arun Daniel, principal consulting engineer for data center and cloud technologies at CDW. The replatforming cut storage needs by up to 80 percent in some cases, and drastically reduced the number of servers required to run AANA’s workloads.
Even after the initial migration, CDW continued to fine-tune AANA’s public cloud environment through its managed services offering. That engagement proved to be vital, as Gilligan says the initial monthly costs for the AWS environment nearly gave her “a heart attack.”
“The more we added, the higher the monthly consumption costs were,” she says. “But CDW was able to make recommendations from one month to the next on how to change that environment to reduce costs.”
CDW continues to manage and monitor the AWS environment, allowing Gilligan and her team to turn their attention to strategic projects that add value to the organization. “We get to focus on our core business,” she says.
EXPLORE: Learn why a multicloud environment might be the best fit for your small business.
Finding the Right Tools for the Job
For Trintech, a financial software provider based in Plano, Texas, the right solution is a mix of on-premises and public cloud resources. The product the company offers to midmarket firms runs on Microsoft Azure. But for its enterprise-class product, Trintech opted for on-premises infrastructure.
“It’s two different product lines,” says Michael Ross, Trintech’s chief product officer. “Midmarket firms tend to want to get up and running really quickly. They’re looking to get a handle on their finances, and they’re fine with a public cloud, multitenant environment. But the larger enterprises have more specific requirements. They like to have their data isolated from other customers’ data, and they want to be able to choose when to perform updates.”
Trintech recently refreshed its on-premises environment with new Dell EMC PowerEdge infrastructure. The refresh featured Dell EMC PowerEdge R740xd and R740xl servers powered by Intel Xeon processors, Dell EMC vSAN Ready Nodes and VMware Cloud.
Click the banner below to receive exclusive cloud content when you register as an Insider.
The new hardware provided the organization with the same compute power for 30 percent less money than its previous infrastructure, Ross says. Trintech’s initial goal was to increase availability from 99.7 percent to 99.95 percent — but the company ended up exceeding that target, and now averages 99.99 percent availability.
“We’re very pleased,” Ross says. “We were able to maintain or improve the user experience, and also decrease our costs per customer.”
Another benefit of the Dell EMC PowerEdge and VMware infrastructure: It mimics the simplicity of the public cloud. “We’re able to fully virtualize the network,” Ross says. “The network switches, routing capabilities, load balancing, all of that used to be separate physical devices. Now those all run within the virtualization stack.”
He adds, “We saw the simplicity of adding compute go up dramatically.”
Photography by Michael Starghill
