Disaster Recovery At Boston College
What would happen with BC’s data and technology in the wake of a disaster? The Blizzard of ’78 paralyzed the region just a few decades ago, so prolonged periods with no power or Internet access is certainly a possibility in New England. Less widespread yet equally impactful incidents such as a fire or flood in BC’s Data Center could also cut off communications and access to critical data.
Groups within ITS at Boston College had been wrestling with this problem, and when Hurricane Katrina struck in 2006, there was renewed interest and urgency.
BC formed a Business Continuity Committee, including representatives from Student Services, Instructional Design, University Libraries, Human Resources, FVP, and Facilities. This group selected the core applications, such as WebCT/Vista, the Portal, and the Library’s Exlibrus system, that would need to be recovered within three days of a disaster affecting BC’s data center. In 2007, the Systems Management group was asked to try to recover some key systems from an off-campus location, with testing to begin in 2008.
After these decisions had been made, Systems Management moved on to the next phase – testing the recovery of critical applications.
In January 2008, eleven members of the Systems Management team conducted recovery at IBM’s Sterling Forest, NY business continuity site. Using their off-site backup tapes, the team recovered two infrastructure applications as well as the BC PeopleSoft Financials and Exlibrus.
The success of this effort resulted in a three-year contract with IBM that includes annual testing and provides access to equipment and a network in the event of an actual disaster. The System Management team plans to run increasingly complex tests each year and continues to work with the Business Continuity Committee to plan for and assess the success of recoveries.
Following industry best practices, the group performs a test recovery of the applications in the St. Clement’s Data Center and documents the procedures. These tests have uncovered numerous issues that would have kept various applications from functioning correctly in the event of an actual disaster.
While ITS hopes none of these procedures have to be implemented because of a real disaster, the Systems Management team is working to assure BC will be ready if one does occur.