The University of Miami drives reliability with Microsoft Azure

Reliability in the land of the Hurricanes

Every institution wants to keep its systems up and running. But for the University of Miami, whose venerated athletics teams are called the Hurricanes, dependability takes on even more significance. With thousands of students, faculty, and staff members; leading-edge research; and Division I athletic excellence all counting on fast, reliable technology services, the University must ensure it provides resilient, rock-solid digital assets—even when severe storms come calling. To help protect and maintain accessibility to vital tools and information under any circumstances, the University of Miami found a dependable resource in Microsoft Azure and a valuable guide in the Azure Well-Architected Framework.

Mitigating risk

The University of Miami needed a cloud solution to maintain communications and operations during hurricanes, other rough weather, and associated disruptions such as power outages and damage to physical infrastructure, in addition to increasing reliability overall.

“We wanted to guard against natural disasters like hurricanes but also other interruptions,” says Mari Lovo, Director of IT, Cloud Infrastructure Services at the University of Miami. “For example, we had a fire near campus that blew out our power and subsequently blew out one of our main campus datacenters.”

The incident was a litmus test for the University of Miami’s systems, and it highlighted gaps in reliability. System failures have significant downstream effects. Outages prevent students and faculty from accessing the resources and services they need, such as the widely used learning management system Blackboard Learn, vital research and analytics tools like IBM SPSS Statistics, and crucial enterprise systems like Workday. The fire prompted the University to begin shifting its on-premises workloads to Azure, to avoid a single point of failure.

Architecting resilience

The move to Azure made things like disaster recovery and high availability easier, and it provided a glimpse of the immense benefits the University could derive from moving other workloads to the cloud platform. The value gained in moving key digital assets, like its authentication services and its public-facing website, led the University of Miami to commit more fully to the cloud for operating reliable, secure workloads that are cost optimized.

The University is now architecting redundant infrastructure for each of its applications. For example, it set up a redundant leg for its Domain Name System (DNS) servers in Azure using Azure DNS. This pivotal move enhanced resilience because crucial applications like email, Workday, and the website were all funneled through its DNS servers. The University of Miami also established a redundant version of the load-balancing systems for its email servers and authentication processes, which gives it global site load balancing between on-premises datacenters and Azure regions.

The University’s IT team built an architecture that provides a redundant path for every digital asset so the University is ready for anything. Its new architecture includes redundant local datacenters, redundant cloud datacenters, and VMs supported by Azure availability zones in both the East US and West US regions of Azure, so that everything was built with redundant versions of Azure ExpressRoute and redundant providers on separate routers.

Delivering dependability

The University of Miami helps ensure reliability by toggling between availability zones to manage traffic and balance loads during upgrades and other planned disruptions, flipping from one zone to another to allow systems to continue running. The availability zones also play a crucial role in the case of unexpected outages—like those from a natural disaster. The zones are equally integral to the University of Miami’s ability to conduct system optimization and maintenance.

“Hurricanes aside, the majority of our outages are planned to allow for maintenance on the machines and upgrades to software,” says Michael Lecuona, Systems Architect at the University of Miami. “With availability zones, we can have a stack running while we’re working on another, without service disruption.”

Enabling new possibilities

The University of Miami not only gained resilience with Azure, but it also developed greater agility and now manages to do more at lower costs. With Azure Automation, for example, the University has drastically reduced the time it takes to clean and clear its C-drive. Plus, that reclaimed drive space prevents additional outages that result from space-dependent server stoppages.

“We used to have to manually manage the C-drive to get space back, and servers crashed if we didn’t,” says Lecuona. “By identifying short-term and disposable folders and implementing automation, we now save hundreds of hours of cleaning time.”

This agility extends to new capabilities, fast deployment of smarter solutions, and better business stakeholder support. For instance, when business stakeholders asked for a data warehouse a few years ago, the University would have needed specific types of drives and servers to fulfill their request, which could reach costs of around $2 million. On Azure today, the University can satisfy this requirement immediately on a pay-per-use structure and have the benefit of scalable storage. Moreover, the University realizes even greater savings by incorporating other Azure features and services such as server reservations into its projects—and achieves the added benefit of real resiliency.

“Having that level of agility is huge,” says Lovo. “We can now skip lengthy processes like procurement, delivery of hardware, and setup time, which cuts our wait time from many months to weeks.”

This cost-effective agility has certainly been an asset during COVID-19. At the outset of the pandemic, the University of Miami’s VPN lacked the licensing capability to immediately support everyone. However, its IT team was able to quickly spin up some virtual machines on Azure to facilitate usage of in-network resources and then have Azure carry the load and leave the University’s on-premises assets unstrained.

Its new, more resilient architecture gives the IT team freedom to innovate as well. “We’re more willing to try things because we don’t have to commit to any hefty hardware investments,” says Lecuona. “We can just spin it up and try it without the risk.”

More and more, the IT team looks to Azure to find creative and reliable solutions to the business requests and challenges that come its way. “Whenever we think of a solution, we think, ‘How can we do this in the cloud versus on-premises?’,” says Lovo. “It not only makes us more resilient but more flexible and nimble as well.”

Real reliability

Hurricane seasons come and go, but the well-architected reliability the University of Miami has built is here to last. This equals peace of mind for Lovo, Lecuona, and their teams. “It’s very nice to know that we have an out-of-region infrastructure that we can use to replicate and protect our servers,” notes Lovo.

Others around the University can continue working and fulfilling the institution’s mission of improving lives—without worrying about the reliability of the tools they depend on. Lecuona concludes, “If our faculty, students, and staff have no comment for us, that’s a win because it means everything is working—they have no issues and nothing to think about beyond their work. We’ve put the technology in place to make their lives easier.”

 Find out more about the University of Miami on Twitter, Facebook, and LinkedIn.