Iowa Health System uses 'cloud' for disaster recovery to survive flood

Iowa Health System survived a flood in 2008 without any downtime to mission-critical systems thanks to archiving system based on object-based software spanning two data centers.

The Iowa Health System didn't set out to design a private cloud, but its unintentional internal storage cloud built with Bycast Inc. software and IBM Corp. and NetApp storage ended up saving its mission-critical data from getting lost in the Iowa flood last year.

The Iowa Health System includes with 26 hospitals and more than 140 clinics or group practices. IHS has three data centers, connected by a 3,200-mile fiber optic network. IT director Tony Langenstein said Bycast StorageGrid software powers a private cloud on IBM/NetApp storage across two data centers, running applications in the hospitals and clinics across the state.

Langenstein said the medical network has about 140 TB of data on IBM TotalStorage DS4100 SAN and IBM-branded NetApp network-attached storage (NAS) -- 70 TB in each data center. About 34 TB is on the Bycast cloud. IHS uses the private cloud for mission-critical systems that store PACS and electronic medical record (EMR) data.

More on cloud disaster recovery
Users go for Simply Continuous cloud disaster recovery service

SunGard plans cloud-based disaster recovery for VMware 

NASA to send lunar data to Nirvanix's cloud for disaster recovery 
Bycast's object-based StorageGrid software is frequently mentioned as a building block for cloud storage services, but when IHS was looking for an archiving system for its PACS and other medical imaging applications when it discovered Bycast in 2005. "We had a SAN, but no tiering of data," Langenstein said. "All of our expensive data was on our expensive SAN without being archived."

Langenstein said IHS chose Bycast because of its close association with IBM (IBM and Bycast have an OEM deal), it was geared for medical systems, and its scalability. "Each storage node is 3 TB, and I can add as many nodes as I want," Langenstein said.

But the key factor was reliability. "Our cloud supports four mission-critical systems," Langenstein said. "The goal is no downtime. If it's not available, it's useless."

The goal of no downtime stood even when the flood hit in June of 2008. Langenstein said the IHS team began putting its disaster recovery (DR) plan into action on June 10, one day after the Cedar River began flooding over. IHS opened its command center in Cedar Falls on June 11, and evacuated the Cedar River data center on June 12 -- a day before the river crested to more than 31 feet.

As a result of the flood, 78 Iowa counties declared disaster, more than 4,000 homes were damaged or destroyed and hundreds of businesses closed. Langenstein said IHS lost a total of one server while moving.

IHS survived by switching over from its Cedar Rapids data center to the Cedar Falls data center about 60 miles away. Langenstein said 130 staffers had to relocate during the disaster recovery process. The staff moved 200 servers between data centers, but completed the task without downtime to any of the mission critical systems, he said.

"We had to do a lot of work to get everything else back and running. But we got the data center switched over with no downtime for users," Langenstein said.

The private cloud continues to play a key role for IHS. The medical network is migrating to a new PACS system, which should accelerate its data growth. Now Langenstein is preparing for a new flood – of data.

"We're already growing more than 50% a year in storage and hardware, just from PACS," Langenstein said. "Now we're adding cardio PACS. I'm scared to see what's going to happen."

Dig Deeper on Disaster recovery storage

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.