When Hurricane Florence struck the southwestern US in 2018 it rained over 14,000 square miles and was considered the nation’s second-rainiest storm in 70 years.
The hurricane struck a chord at many companies, especially Spirit Airlines, which is based in Miramar, Fla., the heart of hurricane country.
SEE: Disaster recovery and business continuity plan (Tech Pro Research)
“I can’t tell you how many times we’ve been asked why we would want to have a data center in Florida,” said Martin Painter, senior manager of IT systems infrastructure at Spirit Airlines. “I would answer that we might experience hurricanes, but there are many other natural disaster events that we don’t experience.”
Spirit also maintains a data center in Detroit. “This enables us to have a DR [disaster recovery] alternative that is outside of the area, but still maintain our data center and legacy applications close to home,” said Painter.
In the event of a hurricane, Painter said, it is necessary to move more than systems. “You also have to move people to the alternate site,” he said. To prepare for disasters, the company has taken steps to improve its disaster recovery response.
For example, during the 2009-2010 timeframe, Spirit improved its DR by using data replication and new data recovery tools, although the resulting DR operation was still heavily hands-on and took six hours. “Since that time, we’ve improved our DR further. We can now get systems up in a three-hour timeframe,” said Painter.
What enabled Spirit Airlines to significantly improve its disaster recovery? See below the following seven solutions Spirit Airlines implemented.
1. Hire top-notch DR talent
“Our CIO made the decision to build our internal talent pool by bringing in other veteran IT engineers. Getting the right people in place to handle a DR was extremely important,” said Painter.
2. Partner with a vendor who specializes in DR
Painter said that Spirit found a vendor (Zerto) that focused solely on DR. “They were there to work with us collaboratively and as a partner during any DR situation,” he added.
3. Have DR-ready operations
“When you’re an airline, you have to think not only about the things going on at the ground level, but about the routes you have to fly,” said Painter. “Our operation folks are well versed in handling disaster scenarios. They make sure that the planes can get from point A to point B, and we have many analytics databases and web servers that support them.”
SEE: Disaster recovery: Tech tips and leadership advice (TechRepublic on Flipboard)
4. Pay special attention to legacy systems
“We have several proprietary databases and also vendors like Microsoft and Oracle. These are legacy systems, and we know that legacy systems don’t necessarily move well with replication technology. That is why we went to a co-location strategy with another data center,” said Painter.
5. Regularly test DR with IT and end users
“We involve users from every business area for major DR testing,” said Painter. “We test everything–servers, logins, phones, PCs, etc. We perform the testing by working through test scenario scripts that we design.”
6. Invest into bandwidth
“We have a one gig burstable pipe between our data centers, which has really benefited our recovery efforts,” said Painter. “Back in the days when it took us eight hours to recover we used 100 megabit pipes.”
7. Testing documentation
“We test our DR documentation as well as our IT when we perform DR testing,” said Painter. “If the documentation isn’t clear, we fix it.”
- Why 23% of companies never test their disaster recovery plan, despite major risks (TechRepublic)
- 4 communication fundamentals that should be in every disaster recovery plan (TechRepublic)
- Cloud sync vs backup: Which disaster recovery works better for business continuity? (ZDNet)
- Data backups: A cheat sheet for professionals (TechRepublic)
- Levels of availability and disaster recovery (ZDNet)
- 5 best practices for disaster management (TechRepublic)