IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.
Sponsor Content
What does this mean?

Moving Off Campus: Disaster Recovery in the Cloud

shutterstock_cloud marketplace

How a two-year institution, Allan Hancock College, is ensuring resilience and increasing uptime by modernizing mission-critical systems.

For many years, a central component of Allan Hancock College’s disaster recovery planning was ensuring mission-critical data was backed up and readily available to be restored. But a shift in payroll procedures changed how the California community college approaches data recovery, resulting in a cloud-based disaster recovery system managed by staff in the Amazon Web Services (AWS) Cloud that has accelerated efforts to modernize critical systems. “A lot of people are intimidated by the cloud environment and think it’s a whole different world,” says Andy Specht, the college’s director of information technology services. “It’s really not that different.”

The challenge

Named after a local community leader, Allan Hancock College is a public community college on California’s Central Coast that serves more than 11,500 students across four campuses each semester. For the college’s Information Technology Services (ITS) department, disaster recovery planning had focused on ensuring mission-critical systems and data were regularly backed up with copies available on site and in other locations for resiliency.

“The paradigm for disaster recovery was having backups of data — could you find and restore whatever critical data you have in your system,” Specht says.

However, when payroll operations transitioned from the county government to the college, that paradigm had to evolve. Regulatory demands required Allan Hancock to not only ensure payroll data was backed up, but that payroll processing systems could keep running with minimal disruption, even in circumstances involving natural or manmade disasters.

Like many higher education institutions, Allan Hancock continues to manage many of its business processes through a legacy on-premises enterprise resource planning (ERP) system. Specht says the college’s Ellucian Banner ERP developed over time into “a sprawling ecosystem” as the institution added and modified functions, including payroll. The legacy system complicated plans for disaster recovery and continuity of services, according to Hancock.

“It’s not optimized for moving around,” he says. “It was designed to be on premises, and you build on it over the years and never take anything out.”

Weighing the options

Working with a consulting group, Allan Hancock considered several options to ensure payroll systems could continue operating in the event of a disaster.

The first option involved building a second on-premises data center at the college’s Lompoc Valley campus, about 20 miles away. But hardware and infrastructure requirements would have easily put costs in the six-figure range. Running payroll systems from the secondary data center would also be dependent on internet connectivity between the two campuses, which could be interrupted by a natural disaster impacting both campuses.

The college determined the solution would be “good enough to store server backups, but not much else,” Specht says. “That was nowhere close enough to having parallel instances of our systems.”

Allan Hancock then considered colocation in a rented space in a commercial data center. But this would have required additional equipment and introduced challenges of managing hardware and systems remotely. The college also explored fully outsourcing the process through a managed disaster recovery provider, known as disaster recovery-as-a-service, but costs were significantly higher. “We didn’t find a proposal we liked,” Specht says.

The final option involved the college’s ITS department managing disaster recovery on its own in the cloud. Doing so would provide greater physical distance for data backups, as well as the capability to stage servers remotely without the need to purchase and manage additional hardware in a remote location. Allan Hancock selected AWS as its cloud provider and staff began building the college’s disaster recovery solution in September 2020.

Building a cloud environment for disaster recovery

The process of moving backup and recovery operations to the cloud took less than two months. The first steps involved establishing an account with AWS and configuring the network. Understanding account structures within cloud service providers like AWS represented the biggest perceived — but ultimately manageable — learning curve for ITS staff, according to Specht.

“The account structure in AWS is very straightforward,” he says. Once you have your organizations set up, it’s not all that different from managing virtual machines. That part is very intuitive.”

The more challenging part of the process involved determining what functions and systems should be shifted to cloud-based disaster recovery. Ultimately the college opted to move about a dozen mission-critical servers, focused primarily on the minimum services needed to run its payroll operations and print checks. It also shifted back-office student registration functions, including financial aid and processing, to cloud-based backup and recovery.

“The AWS failover is for key systems when you need to have them back in hours instead of days,” Specht says. “We tried to make things a little easier by not bringing over servers that aren’t essential or that were very large.”

For example, the college opted to leave behind document management systems for viewing attachments and other less critical functions of the ERP system, which will continue to be backed up and stored in multiple locations.

From there, Allan Hancock began copying data from its on-premises data center to the AWS Cloud. To facilitate the process, the college opted to use CloudEndure Disaster Recovery, business continuity software offered by AWS, which can replicate entire servers, including operating systems, system stage configurations, databases, applications, and files. CloudEndure Disaster Recovery was selected for its ability to replicate a wide range of applications from on-premises data centers to AWS and enabling them to run on AWS just as they do on-premises.

It took just six weeks for Allan Hancock staff to replicate their servers to the cloud.

With the cloud-based recovery system up and running, in the case of emergency, ITS staff can use CloudEndure Disaster Recovery to spin up the college’s virtual servers in AWS in their fully provisioned state within minutes. Once configuration changes are made to the college’s domain name server (DNS), live services could be ready for use within just a few hours, according to Specht.

Fringe benefits — and a foundation for future growth

Allan Hancock College hasn’t had to use its new failover system in a real emergency yet, but ITS staff members are already seeing significant improvements in disaster recovery procedures — as well as what Specht sees as fringe benefits.

Instead of backing up servers at set intervals, CloudEndure Disaster Recovery continuously replicates servers and data in the cloud. “We now have snapshots at a minimum of every five minutes, so if we have a disaster, that’s very minimal data loss,” Specht says. “That’s a huge advantage over other processes where we have snapshots once or twice a day.”

The cloud-based approach is also helping reduce on-premises dependencies that could complicate disaster recovery. For example, the college has moved its on-premises DNS to AWS.“

The primary benefit is that it’s somewhat easier to manage and have everything in one spot,” Specht says. “It also simplifies disaster recovery — we’d have to set up a DNS anyway, and this allows us to avoid having to copy over and configure it in the cloud.”

What’s next with the cloud

Allan Hancock’s success moving disaster recovery to the cloud is prompting the college to examine a more difficult choice — what to do with its legacy ERP system. Staff are evaluating a range of options, including replacing the system with a native cloud-based solution or migrating the existing one into a cloud environment.

“In the end, Ellucian Banner will leave the on-premises environment, but we’re not sure what that will look like,” Specht says. He credits their learnings from modernizing their disaster recovery processes with accelerating their shift to cloud migration.

The college’s ITS staff is also becoming more comfortable with AWS and cloud methodologies. Our developers are now experimenting with new web applications in AWS that use Amazon’s Simple Email Service (SES). “One advantage I like is my staff growing their skillsets,” Specht says.