4. Develop Specific Risk Scenarios For Your Exercises
Many enterprises conduct their DR exercises without specific scenarios; they tell the response team to assume the data center is "a smoking hole." It is important, however, to define specific risk scenarios even for DR testing for two main reasons: 1) It provides a more realistic situation for the response team to react to, and 2) different scenarios require different actions from the IT staff. For example, the DR plan for a short outage at the primary data center that only requires resuming operations would be different from a long-term outage that requires failover (and eventually failback), which in turn would be different from scenarios where only portions of the IT infrastructure were down.
5. Run Joint Exercises With Business Continuity (BC) Teams
In our research, Forrester found that many BC and DR teams run all of their exercises separately and often fail even to communicate when they run exercises. However, you should aim to exercise the full enterprise BC and DR concurrently at least once per year. This is especially important if the data center is in the same location as corporate headquarters.
6. Vary Exercise Types From Technical Tests to Walk-Throughs
A common misconception in IT is that walk-throughs and tabletop exercises are not necessary for DR exercises. While it's true that these types of exercises won't test the technical capabilities of a failover, they are still critical for training, awareness, and preparedness. Interviewees told us that the majority of the time, exercises that didn't go as planned actually struggled most with communication and employees' understanding of their roles during the exercise. Non-technical exercises such as walk-throughs and tabletops will help make these processes go more smoothly.
7. Make Sure to Test All IT Infrastructure Concurrently at Least Once Per Year
Waiting longer than a year risks too much change in IT environments and personnel -- you need to bring new staff members throughout the organization up to speed on DR plans. The most advanced firms run full DR tests as often as four times per year. In between full tests, most firms conduct component tests that vary in frequency depending on the criticality of the systems and rate of change in the environment.
8. Identify Members for the Core DR Response Team
The stress of working under time and resource restraints for long hours, often during nights and weekends, is something people cope with in different manners. When picking a core response team to lead IT recovery, it's important to pick people who can work under extreme amounts of pressure (and sleep deprivation). During an exercise or test, identify those individuals who can remain calm and collected.
Sign up for CIO Asia eNewsletters.