With a staff of 22, the IT department at Manhattan College resources are quite limited. “Like a Swiss Army knife, we can cover lots of things, but we don't have the expertise in any one area,” says Jake Holmquist, the CIO at Manhattan College. “Much of our staff is made up of IT generalists, so it’s difficult to make a good solution great.” The college had been using many on premise tools but as needs grew and issues became more complicated and costly.
Manhattan College improves information technology services to better support students and faculty
With its small IT staff, Manhattan College had no economies of scale for larger projects, which proved both inefficient and expensive. By shifting to Google Cloud, the staff now spends less time in the datacenter and more time devoted to the people they serve.
"Our journey has been all about breaking down cloud barriers, and becoming more comfortable about a concept that was initially foreign to both users and IT staff. As we broke down each barrier, we made it easier to take the next steps."Jake Holmquist, CIO, Manhattan College
Shifting to the cloud
“We turned to the cloud,” says Holmquist. “In most cases, it has been the best answer.” Since 2008, Manhattan College has used Google tools. First came the transition to Gmail, and that, says Holmquist, “was the foot in the door that we in IT needed to show the rest of campus that it was okay to operate in the cloud.” Both from what he calls a trust standpoint and familiarity and ease with Google tools, it made sense to stick with Google. “Having existing Google Workspace accounts really made it easy for our IT team to get access to Google Cloud tools, and to delegate access once we spun up services,” he says.
The IT team’s approach started with low-risk redundant and "dataless" systems: DNS, and small websites with no personal information. “As we accelerated our cloud-adoption and migration of more user-facing services,” says Holmquist, “we started inserting ‘secure cloud environment’ into our messaging about upgrades and migrations. Along with that messaging included a new layer of redundancy and high-availability services performed better, were more redundant and more reliable once migrated to the cloud—something that we could never have achieved on campus.”
Over time, “Our journey has been all about breaking down cloud barriers, and becoming more comfortable about a concept that was initially foreign to both users and IT staff,” he says. “As we broke down each barrier, we made it easier to take the next steps.”
Holmquist notes that when they first adopted Google Cloud, in 2014, the tools were not well documented. Since then, however, “the tools are much easier to use, more can be done through the UI, and documentation is far better with more useful examples and a larger community of support.”
Challenges prove rewarding
Manhattan College’s primary challenge involved the big commercial applications they had migrated. Most were designed for the datacenter—and, says Holmquist, “in our experience, have not performed well ‘out of the box’ when moved to the cloud. Our biggest hurdle has been the initial configuration of these (not well-documented) settings to optimize the applications for the new opportunities that the cloud provides.”
Yet he adds that their greatest achievement to date actually resulted from that challenge: In July 2017, they implemented ‘Banner 9,’ an upgrade to their prior system.
“A typical deployment in our datacenter meant a six-figure hardware purchase that we were not guaranteed to be delivered and provisioned in time for ample testing and Summer 2017 go-live,” he says. “Instead, we took the unprecedented approach of deploying these new Banner 9 components in Google Cloud’s Compute Engine. We were able to quickly and easily spin up various components during the installation and upgrade testing. When we failed, we quickly deleted the instance and started over in a matter of minutes. With the help of Google Support, we were able to identify a number of settings related to site-to-site VPN (Banner 9 components in Google Cloud talking back to on-premise database), application server performance tuning, and finally high-availability and SSL.”
In the end, they were able to deploy a production environment with “excellent performance and a level of high-availability that we could not have achieved on campus.”
Now the college can spin up and provision a new system in a matter of minutes. “If we need to start over, we can delete it even faster,” he says. “Better yet, more of our staff can spin up services as needed, instead of waiting for a system administrator to provision for them.”
Savoring progress and looking ahead
Above all, Holmquist has aimed to get the IT team out of the datacenter and more involved with user-facing initiatives. “We've certainly crossed a threshold, in that more of our systems are hosted in Google Cloud than in our datacenter,” he says. “Instead of maintaining servers, replacing failed components, and applying patches, we are now focusing on making our applications run more efficiently which results in a more measurable benefit to our end-users.”
Additionally, features that campus users now rely upon—such as excellent performance and high-availability—were once far too costly for the college to implement on its own, especially with its limited resources. “With Google Cloud, these features are available to us without huge upfront costs, and, in most cases, are just part of the toolset,” he says. “Best of all, the costs are a fraction of what we would have paid on campus and all something that can be managed effectively with existing staff.”
Much progress has been made at Manhattan College, yet more plans are in the works. The school will continue replacing servers in the datacenter through attrition, and aims to explore better, more efficient ways to deliver services in the cloud.
Holmquist is pleased to have freed up IT staff to focus more resources on "business process improvement" through electronic workflows, and to make better sense of campus data through data reporting and analytics initiatives with campus offices. The department has successfully transformed a traditionally internal-facing IT position into “a user-facing technology trainer to empower users with these new tools.” He is eager to see how Machine Learning and Analytic tools might be used to support this effort.
For now, says Holmquist, “Google Cloud has given us a toolset to solve complex problems that we could not have solved on our own.”