Summary

---

Developing a Test Plan

What good is a DR plan if it is not tested? Not very good-especially since the whole purpose for putting together a DR plan is to be prepared in the event you suffer any measured loss of data, equipment or services. This plan is supposed to give us an edge on such events, but if we do not test it, our planned outcome may be entirely different. Therefore, take this very simple methodology to heart as you begin this process: Plan your test, then test your plan. This very simple statement means that you must define a scope for your test because it is within this scope that you will find success or failure of your DR plan. If you fail to do this, you lose consistency and repeatability of your DR plan. Remember, we are testing the DR plan so we are prepared; failure during a test that is documented and addressed is actually a success. However, failures that are ignored remain failures.

Planning the Test

When you begin the process of planning your test, start at the result and work backward, much like we described at the beginning of this appendix. If you can define your results as goals, you should be able to build the steps necessary to achieve each goal. For example, the end result of your DR plan is that you would have your Online Reservations System (ORS) up and available within 14 hours. To accomplish that in the face of a declared disaster, you need to take the following steps:

1. Disaster is declared: Assumption-Facility inaccessible. To successfully test, you must make assumptions in order to test each level of the response teams. Other examples are as follows:

   1. System failure

   2. Application corruption

   3. Advancing weather conditions

2. Contact response teams, including

   1. Management team lead

   2. Emergency on-call team lead

   3. Assessment team lead

   4. Recovery team lead

3. Hot-site facility must be notified if this is required.

4. Off-site vendor must be notified and most recent tapes need to be delivered to hot-site facility.

5. Install/configure backup server.

6. Recover backup server's database.

7. ORS recovery machine is installed/configured with an OS and backup client software.

8. Import/inventory off-site tapes.

9. Begin recovery of data.

In this example, each step may have substantially more detail that can be expanded upon, but it gives you a good idea of where you need to go with planning your test. Again, the more detailed your test plan, the greater the probability is that you will find success during its execution. Do not let verbosity be the guiding light for whether or not a test plan is adequate; begin where you are with what you have and use the results of the tests, including both failures and successes, to build a stronger, more resilient plan. You may want to test only certain components of your disaster recovery plan, such as the restore component. In that case, make your test plan modular; in other words, create it in such a way that you can pull out certain components without compromising consistency. We strongly recommend that you perform these tests on your off-site media at least every six months- preferably quarterly, if you are allowed. Just remember to stay within the scope of your test plan as you are executing the test plan, so you can see the holes. Do not stop, rather, continue until you can go no further. Then document everything so you can plan your test better the next time.

---