The Changing Face of Disaster Recovery
SHARE recently announced a list of five disaster recovery and business continuity recommendations that it says all mainframe shops would do well to follow.
Disaster recovery (DR) and business continuity (BC) planning have been going on for about as long as companies have been running mainframes.
The rub—according to some industry watchers—is that DR and BC in the post-9/11 and post-Sarbanes-Oxley era create a whole new ballgame.
With this as its starting point, the SHARE user group recently announced a list of five DR and BC recommendations it says all mainframe shops would do well to follow.
SHARE's list notes what businesses should employ to ensure business continuity. First and foremost, organizations should create what SHARE calls a "multiple footprint" for data recovery. This means establishing and maintaining more than one data recovery site—preferably in geographically dispersed locales.
In fact, some mainframe technologists say, geographic dispersal is nothing more than common sense. "Something that 9/11 did bring home is that true disasters can be regional in scope. I think the main DR focus on 9/10 was a fire destroying a single processing center," says Kevin Kinney, a mainframe operator with a prominent financial services firm based in the Northeast. "On 9/11, several companies near the [World Trade Center] pulled the ripcord and all of a sudden Sungard in [Philadelphia] became a very busy place."
As a result, Kinney says, DR planners had to take a new (and hitherto undreamt of) scenario into their BC planning considerations. "Something that DR planners now have to ask themselves is: ‘If my entire city had to declare an emergency, would our DR provider have enough capacity to service us?’ That’s the sort of question that can really ruin your day," he concludes.
Elsewhere, SHARE counsels, organizations need to practice predictive monitoring—knowing in advance whether transactions are at risk of not occurring in timely fashion. In any case, companies should take pre-emptive action to correct faults and prevent potentially adverse outcomes. In addition, SHARE says, organizations should consider employing autonomic computing. In the event of a calamity, SHARE representatives explain, self-diagnosing and self-healing systems and applications are one way to guarantee that critical business functions and services will continue to be available.
Just as important, mainframe shops need to provide remote access for staff—including connectivity from mobile devices. This helps ensure both application and personnel availability, as well as effective geographic dispersal, SHARE officials say.
Finally, the user group concludes, organizations need to deliver 24x7 IT support. This should be a fait accompli in many mainframe environments, in any case.
Stephen Swoyer is a Nashville, TN-based freelance journalist who writes about technology.