Disaster recovery tips and lessons learned

In his keynote address at the recent Network World IT Roadmap event in the US, Marshall Lancaster, vice president of IT, Enterprise Infrastructure Services for United Stationers, offered up lots of tips and lessons learned that can help companies implement effective -- and cost-effective -- disaster recovery plans (Read our main story on his experiences here.)

Have a backup communications plan in place: Following Hurricane Katrina, mobile phones were essentially useless because they were all homed out of a New Orleans central office that was out of commission. "You could call out, but you couldn't call in. So we were able to call people, but not each other," Lancaster said. Lagasse ended up relying on its internal associate Web site to instruct employees to call an existing toll-free employee hotline in order to get in touch with them.

Be wary of tape recovery: "If you're going to use [tape] for anything you really care about, just make sure you understand it very well because it's complex and you need to understand what occurs if there is a problem," Lagasse said. It's best to rely on a premier storage provider for Tier 1 and 2 applications; use tape only for Tier 3.

Be aware of force majeure clauses: Many vendor contracts have such clauses that essentially say in the event of certain major disasters, they are not required to provide assistance -- so don't count on them.

Take advantage of technology refresh cycles: As a way to keep costs down, move older equipment into your backup environment as you replace it.

Use 'owned space': Lagasse's backup data center was housed in its parent company's data center, another cost-saving technique vs. leasing backup space.

Don't try to protect against everything: Lagasse performed scenario pre-planning to determine which events were most likely to occur at its New Orleans headquarters and what the impact would be. Hurricanes were obviously high on the list and, thus, the company focused its resources in protecting against them.

Consolidate applications and databases: Running multiple applications and databases on a single server is another way to keep costs down in your disaster recovery environment. "Performance requirements aren't nearly as high as you expect when a disaster recovery occurs," Lancaster noted.

• Bookmark this page
• Share this article
  •  
  • facebook
  • slashdot
  • digg
  • Reddit
  • stumbleupon
  • linkedin
  • twitter
• Got more on this story? Email Computerworld
• Follow Computerworld on twitter

• Share
  Share this article

  •  google+
  • facebookfacebook
  • slashdotslashdot
  • diggdigg
  • RedditReddit
  • stumbleuponstumbleupon
  • linkedinlinkedin
  • twittertwitter
• print
• email
• Bookmark