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

Lessons learned from 9-11: Disaster recovery dos and don'ts

At Oracle OpenWorld, Veritas CEO Gary Bloom described his company's post-September 11 data recovery mission and offered disaster recovery preparedness advice.

The September 11th terrorist attacks taught many businesses the sobering lesson that their data recovery strategies are woefully inadequate, said Gary Bloom, CEO of Veritas Software Corp., a storage systems vendor based in Mountain View, CA.

In a keynote speech at Oracle OpenWorld 2001 in San Francisco on Dec. 5, Bloom described Veritas' emergency response to the attacks and offered advice for data recovery preparedness.

The scenario of having about 100 Veritas customers lose major data repositories "on the same day at the same time was unthinkable," Bloom said. When the unthinkable happened, "we as a company had to make a quick shift because we had a job to do."

A team of 650 Veritas employees set to work restoring a massive volume of information," said Bloom. Their work resulted in recovery of 100 percent of the 100 companies' 35 terabytes of data that had been backed up. The only unrecovered data was data that had not been backed up.

Data recovery dos and don'ts

In his Oracle OpenWorld presentation, Bloom listed some of the disaster recovery lessons highlighted by the Sept. 11th attacks. In addition, this list of dos and don'ts includes tips offered by Veritas in its disaster recovery seminars.

Don't place back up facilities near each other. "The location of back up sites matters," said Bloom. A World Trade Center customer, for example, had a back up facility located several blocks away from its WTC offices. Data was lost at both sites.

Businesses have traditionally put back up facilities close together because of the technology limitations of hardware-based storage systems, Bloom said. Today, the availability of clustering and online storage technologies makes it possible to place data in repositories all over the world.

Do regular backups of data residing on desktops and laptops. "Synchronize desktop and laptop data to the server daily to be totally protected," Bloom said.

Don't just back up data. Many companies backed up data, but they didn't back up data catalogs and directories, Bloom said. As a result, data recovery was slowed down by the lack of a means to organize data and get the appropriate permissions for access to data.

Do run similarly designed back up and storage environments, so that incompatibility issues won't derail data recovery. "Make your life easier by having heterogeneous software that works the same from environment to environment," Bloom said.

Do make sure that there are back up facilities have the hardware, software, network connectivity, and services needed to run all operations. Until Sept. 11, most companies did not plan for the possibility that an entire facility would be completely wiped out. Unfortunately, they now have to plan for that possibility.

Do be prepared to re-establish systems management capabilities quickly. Having back up management capabilities -- such as monitoring, job execution, and security features -- will make it easier to establish a data center command center in a new location.

Do prepare by training internal disaster response teams and identifying the applications and business operations which should receive first priority in a recovery. A well-trained and well-rehearsed crisis response team will be able to restore data and bring operations back up quickly.

Attendance increases at Veritas' disaster recovery seminars show that the recent tragedy has made disaster recovery planning a priority for businesses, Bloom said. The seminars used to pull in 75 to 100 people. After Sept. 11, 300 to 400 people have attended.

Dig Deeper on Windows Server storage management

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.