Phase 3 — Evaluating the disaster recovery

The final phase of disaster recovery is evaluation. You need to determine whether the disaster recovery was successful. If it was not successful, you need to determine the reasons for the failure.

Obviously, evaluation is most effective in the context of a simulation; if you find a problem, you can correct it before you run the next simulation. And you can continue to run simulations until you get the results you want. Therefore, by carefully examining the results of your disaster recovery simulations, along with making adjustments in the preparation and performance phases of disaster recovery, you can ensure successful recovery when a real disaster strikes.

You may ask the following questions to evaluate the disaster recovery:

• Were all necessary databases recovered, and were they recovered correctly (without losing too many in-flight transactions)? If not, why did the error occur? The following problems may have caused the error:
  • Incorrect or incomplete preparation at the home site
  • Unexpected variations at the disaster recovery site
• Was performance acceptable? Did you meet service-level agreements for the length of the database outages? If not, why? The following problems may have caused the unacceptable performance:
  • Need for more frequent backups
  • Need for more frequent change accumulations
  • Need to improve recovery profile options
  • Inadequate resources (such as tape drives, CPU cycles, and DASD)
  • Need for more recovery points
  • Inappropriate backup stacking strategy
  • Indexes were recovered rather than rebuilt