I have been asked by many businesses to mention the yearly test problems. Here are the 3 major ones:
Yearly test is not reliable: it neutralized from unexpected incidents which accurse in real disaster. Moreover, in real disaster the few IT members or experts for specific applications or hardware may not be available.
It’s a big risk to the organization to check the recoverability once a year. The production site it’s a dynamic environment with many changes, while the secondary site is much more static. Along the year the business is at risk. Moreover, long periods of tests, dose not provides the organization, the confidence and experience necessary to respond to real emergency.
Yearly test is not up-to-date: It is likely that the team putting together the disaster-recovery plan considered some type of priority ranking methodology for the applications and business functions. We have to consider the fact that along the time, the priority ranking change. Acknowledging the fact that criticality ranking of the application is not static, businesses can’t be satisfied with a yearly test. The solution is a weekly or monthly test.