News Stay informed about the latest enterprise technology news and product updates.

Katrina data management snafus compound chaos

A lack of planning and IT skills leaves the Red Cross scrambling to organize data into a system for refugees to search for missing loved ones.

In the weeks that followed Hurricane Katrina, IT professionals and volunteers scrambled to assemble technology to reunite evacuees with their loved ones. But while technology eventually prevailed, many reunions were delayed as IT and bureaucracy clashed in the face of the unprecedented disaster.

On arrival at Houston's Astrodome shortly after the hurricane, approximately 25,000 evacuees filled out a form with their personal information and gave it to the American Red Cross. Volunteers then entered the data into a pilot database program called the Coordinated Assistance Network (CAN), run by a group of nonprofit organizations, including the American Red Cross and the Federal Emergency Management Agency (FEMA).

"CAN is an Internet-based database in which many agencies post information for the purpose of sharing data," said Andrew Adams, an IT production analyst at First Data Corp., which was drafted by the American Red Cross to help with the effort. But a bureaucratic oversight prevented data from being shared. "The problem was that client releases were required if information was to be shared outside of the agency's system and the Red Cross didn't get this release from the victims," Adams said.

In other words, the names of close to 20,000 evacuees were entered into the system before anyone realized that none of the names could be searched by the public. Furthermore, bugs in the application meant that it took several days for the CAN data to be uploaded into the official International Committee for the Red Cross database, which could then be made available to the public.

Separate volunteer efforts sprung up like daisies to get the data online and available. "We created a quick-access database in SQL Server to serve as a temporary repository while the Feds finished CAN," Adams said. At least 20 different search sites appeared on the Web, prompting Yahoo! to build a tool to simultaneously search multiple databases. However, at press time, this search function still didn't work across all of them.

Bill Fitler, a retired IT specialist and volunteer for the American Red Cross, said CAN was never meant to be a public database, although eventually a subset of it was released to public Web sites like "You have to appreciate the balance between confidentiality and the need to search the data. If everyone started searching CAN, it would be totally nonfunctional," he said.

CAN was developed in the wake of the Sept. 11 terrorist attacks and was piloted in six cities, including the District of Columbia, New York City, Oklahoma City, San Francisco, Seattle and New Orleans. It wasn't supposed to be rolled out officially until December, according to Michael Raymond, a data warehousing expert at Reliant Energy Inc., also volunteering for the Red Cross. "There's a team of people still running around the country training the shelters on how to use this system. [The] federal government rushed it out," he said.

Problems tracking people persisted. Technology For All, a nonprofit organization that helps connect community centers to the Internet, said another problem was recording people who left the Astrodome. "When someone finds someone, they tend to run out the door without leaving a forwarding address, so there is no way of connecting with them," said William Reed, president and CEO of Technology For All.

Reed believes there are important data management lessons to be learned from this tragedy. "Our world is an information society. FEMA and the Red Cross haven't seen the value of how we could help people find their loved ones, housing [and] jobs. In disasters in the future, when you bring people into relief centers, IT has to be made a vital part of helping people, not an afterthought. In terms of databases, particularly in a large tragedy, there needs to be one unified data store to help people find each other, so that we don't have to create tools on the fly to search all these different repositories."

"This disaster will force the Red Cross to come kicking and screaming into the 21st century," Raymond added.

It's worth noting that data management snafus aren't unique to emergencies like Katrina. Organizations everywhere struggle to manage databases that are replicated again and again without having the right information. The same is true for redundant file data. The tragedy of this situation is that at a time when technology could have made a huge impact on how quickly the suffering of Katrina's victims was alleviated, the right resources were unavailable.

Another Astrodome volunteer, Rahul Mehta, founder, president and CEO of storage software firm NuView Inc., said it was tragic that volunteers couldn't use technology to unite families more quickly. But "it was also very gratifying as volunteers to see the excitement on people's faces as they were reunited with their families using the Web in spite of all the different databases," he said. "Hopefully, we can solve this problem before the next tragedy."

This story originally appeared on

Dig Deeper on Enterprise data storage management

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.