GlobalNOC on target in XSEDE failover test

It’s 3am – do you know where your data is?

To answer affirmatively, computer systems and services requiring a high degree of availability and reliability must be able to switch over to a backup system. On March 26, the Global Networks Operation Center (GlobalNOC) at IU took part in a backup test for the Extreme Science and Engineering Discovery Environment (XSEDE).

From 9am to 1pm, engineers simulated a complete service failure at the XSEDE operations center (XOC) at the University of Illinois at Urbana-Champaign. The XOC functions as the full-time help desk and call center for the cross-country cyberinfrastructure XSEDE integrates.

“The scientific community within XSEDE cares about the availability of services, and we attempt to enhance that availability by testing failover procedures on the back end,” said Stephen McNally, XSEDE systems operational support group manager.

During the simulated outage, XOC handed off phone calls and tickets to GlobalNOC staff. As planned, the GlobalNOC service desk took over immediately, fielding all routed phone calls and service tickets – and generally operating as the XOC for the duration of the exercise.

“We were very pleased at how smooth the transition went for our first failover,” said XOC manager Mike Pingleton. “I was impressed with the professionalism and helpfulness of the GlobalNOC staff, and I think they were a very good choice for a failover site.”

“The collaboration between the XOC and the Service Desk was great,” agreed Adam Williamson, GlobalNOC service desk supervisor. “Safeguarding against system failures is made possible through collaborations like the one between XSEDE, IU, XOC, and the GlobalNOC Service Desk. We plan on performing more failovers every three months to ensure our operations have experience in case a real event occurs.”

As the GlobalNOC has shown, scholars linked to XSEDE need not worry: If a storm turns out the lights, their work will be safe.