Disaster recovery network pays off in multiple ways

11.04.2006
"When I came to Children's in 2001, I brought the philosophy that if we do things right, then issues like [the Health Insurance Portability and Accountability Act] will take care of themselves," says Albert Oriol, IS program office director and data security officer at The Children's Hospital in Denver, an integrated health delivery system affiliated with the University of Colorado. His philosophy has paid dividends, including cost savings by enabling the hospital to combine disaster recovery and fail-over and simplify its upcoming move to a new state-of-the-art facility seven miles away.

"Most organizations have their fail-over boxes sitting next to their main systems in the data center, and then they have to buy a third set of boxes for their disaster recovery site," Oriol says. "We are able to put our fail-over boxes in the disaster recovery site and eliminate the need for that third set of boxes entirely."

Children's has split its storage clusters, with half at the disaster recovery site and the rest at the main data center downtown. The key to making this work is a very fast, responsive network connection using Ciena Corp.'s Adaptive WAN products running over a private 2Gbit/sec. fiber physical layer. It uses EMC Corp.'s Symmetrix Remote Data Facility and MirrorView to support synchronous backup of all data to the disaster site. The result: "Our tests show that it takes less than five minutes to fail-over, validate and move on," Oriol says.

And because of that, the hospital's disaster recovery operation runs on full duplicates of the computers in its data center rather than minimal systems. "When we do have a cutover, we plan to run on the disaster site until we need to move back to prevent a service interruption, rather than having a second delay to move back to the main data center," Oriol says.

This is particularly important as Children's moves to its new electronic medical record (EMR) for inpatients. The EMR is already in use for outpatients, and the first phase of the rollout in the main hospital is in process now, with the goal of moving all medical charts off paper. Obviously, with medical care dependent on computer access, the hospital cannot afford lengthy downtime.

The system also saves on tape handling. The hospital writes to the tape library at the disaster recovery site directly from its storage-area network disks.