The Department of Information Technology (IT) at the University of British Columbia (UBC) isn’t a typical IT organisation. In addition to providing IT support and services to the UBC community, it also provides infrastructure-as-a-service (IaaS) and disaster recovery (DR) to 26 universities across the province. UBC’s IT infrastructure serves more than 250,000 students and 60,000 faculty and staff throughout Western Canada.
To enable this extraordinary level of IT delivery, UBC developed world-class data centres capable of enterprise business operations akin to those of Fortune 500 companies. Thousands of VMs run critical databases and applications containing several petabytes of data that must be available 24.7.365. As a highly regarded medical research university, one of UBC’s priorities is to protect data containing important cancer and Alzheimer’s research.
As expected, UBC’s Department of IT must deliver the highest level of service. This became a problem during deployment of a legacy backup tool in the VMware vSphere environment. Although the legacy tool had backed up UBC’s physical servers successfully for 15 years, it was ill fated in a virtualised environment.
“The main problem was complexity—both in deployment and use,” said Mario Angers, Senior Manager of Systems at UBC. “Backup jobs either went past our backup window or failed, despite the vendor’s assistance during the yearlong implementation. We even participated in a one-week training seminar hosted by the vendor but felt no closer to understanding the tool’s complexities. We couldn’t ask our customers to use this product for self-service backup and recovery.”
It was challenging for UBC’s Department of IT to offer service level agreements (SLAs) to customers, much less meet recovery time and recovery points objectives (RTPOs).
“After the vendor suggested that we consider restructuring our IT infrastructure to fit their backup strategy; we knew we needed a solution that would align with our vision, not conflict with it.” Angers said. “Our vision is to provide VMDKs for absolutely everything on NFS to gain speed and efficiency, but the legacy backup tool wasn’t capable.”
The legacy tool had additional drawbacks: It didn’t fully integrate with UBC’s NetApp FAS system, and it didn’t provide VMware vCloud Director support. But that wasn’t the only problem. UBC didn’t have full visibility into the virtualized environments that support its own university community and 26 other university communities. Reduced visibility also made capacity planning difficult.
“Without full visibility, we couldn’t proactively identify and resolve issues before customers were impacted. At times, customers told us about outages before we were aware of them”, Angers said. “We spent a lot of time manually monitoring each virtualized environment and dealing with incident-related support requests, which delayed requests that were project-related. And, without full visibility we struggled with capacity and performance management, which play a huge part in ensuring our IaaS is successful.”
Angers said UBC’s customers have the choice to contract with any IT service provider; therefore, simplifying IT for customers and reducing time-to-resolution averages are key competitive advantages.
“From a business perspective, we must deliver IT services that are better, more economical, and more compelling than other providers,” he said.
Veeam® Backup & Replication™ and Veeam Management Pack™ (MP) for System Center met all of UBC’s needs.
Veeam Backup & Replication full integrates with NetApp FAS, enabling the university to establish SLAs for the first time. UBC uses Veeam to make super-fast backups from NetApp storage snapshots with little to no impact on the production environment. Meeting RTPOs is easy because Veeam offers high-speed replication for guaranteed restore points and high-speed recovery features. Instant VM Recovery® restores VMs to the production environment by running them directly from backup files and Veeam Explorer™ for Storage Snapshots recovers individual items or entire VMs quickly and efficiently from NetApp Snapshot, SnapMirror and SnapVault.
“Veeam made it easy for us to operationalize VM backup and recovery,” said Brent Dunington, Systems Architect at UBC. “Veeam is simple to deploy and use, and it integrates so seamlessly with NetApp that meeting RPOs and RTOs became non- issues for us.”
Veeam Backup & Replication also helps UBC deliver IaaS to customers by backing up and restoring VMs managed by vCloud Director.
“We’ve performed successful vApp recovery many times with Veeam,” Dunington said. “Configuration was simple and required little to no training for our operations staff.”
Veeam MP solved UBC’s visibility problem by fully integrating with Microsoft System Center to provide complete visibility of all the virtual and physical environments UBC maintains (it’s own and those of the 26 universities) from the System Center console, enabling proactive management.
“Before Veeam MP, we had no visibility into our virtualised environment because of limited monitoring and reporting,” said Sanja LeBlanc, Manager of Systems at UBC. “We operated in a manual and reactive mode. A customer would call about a problem, we’d figure out what was broken, and we’d fix it, which could take minutes, hours or days.”
UBC was in reactive mode when one of the NetApp storage arrays failed. Veeam MP hadn’t been deployed yet, so no one knew about the failed array until a customer called to report a problem.
“It took several hours to bring the storage array back online,” LeBlanc explained. “Two years later when Veeam MP was monitoring and reporting on our virtual environment, a NetApp controller went offiine, and Veeam MP notified us immediately. We resolved the issue quickly before customers were impacted.”
For any university, a cost-effective solution is a must. Veeam MP demonstrates its value by helping UBC understand growth areas and trends. For example, before purchasing additional compute and storage capacity for itself and the 26 universities, UBC uses Veeam MP for capacity planning and right sizing to reclaim and reallocate underutilized resources. Therefore, UBC maximizes budgets (for itself and the universities) by avoiding over purchasing hardware.
“When we look at the combination of Veeam Backup & Replication and Veeam MP, we see benefits on both the business and the technology sides,” Angers said. “On the business side, Veeam Backup & Replication helps us deliver IT services that are more compelling and costeffective than other providers. Veeam’s support for vCloud Director is critical in delivering cloud services to customers, and because Veeam will be four times less expensive than the legacy backup tool over a five-year period, we’re able to provide affordable IT to customers. Veeam MP is also important because it helps us keep customers happy. Veeam MP gives us much-needed visibility into our virtualised environment, which lets us resolve issues before they affect customers.”
Angers said Veeam’s pricing is easy to understand: Veeam is licensed per CPU socket. “The legacy tool appeared to be licensed based on the temperature outside,” he said. Dunington explained Veeam’s benefits on the technology side.
“Veeam Backup & Replication integrates so tightly with NetApp that we can make fast, frequent backups from snapshots with zero impact on production. That integration also gives us good dedupe and compression ratios across our backup jobs. Dedupe values average between 2.5 and 6 times and compression levels average 2.5 to 4 times.
“Veeam MP provides technology benefits too. It lets us see our virtual and physical environments under a single pane of glass, which gives us unified monitoring, reporting and capacity planning.”
UBC provides IT services to its own massive university community as well as to 26 universities in Western Canada. UBC made a commitment to simplify IaaS and DR for customers but knew from experience that its backup tool was too complex for them to deploy and use.
UBC didn’t have full visibility into its virtualised environment, limiting its ability to identify and resolve issues before they affected customers.