What NOT To Do -

Worst Practices For Disaster Planning and Recovery

Introduction

Natural disasters such as Hurricane Katrina keep network administrators acutely aware of the need for a solid data backup and disaster recovery plan to ensure business continuity. But too many companies observe worst practices, thereby dooming the plan - and their business. This page lists a series of worst practices in the hopes that showing what happens if you do it wrong will illustrate how to do it right. Any of the items in the list below can sink your boat - two or more add up to the end of your business and possibly your career.

Fail to win management support. One of the worst things you can do for just about any IT project is fail to win management support. Without the executives on your side to back you up with the authority and the budget to do the job, the backup and recovery plan is virtually doomed.

Provide no risk assessment. Without some form of risk or impact assessment, it is impossible to know which assets are critical and which ones are expendable. By wasting resources protecting expendable assets while leaving critical assets out of the plan, you will make the plan a failure.

No written plan. People come and go. Maybe the parties that were present when the backup and disaster recovery plan was devised were so brilliant they didn't need to write it down. But, if you do not write down your well-defined and clearly documented plan so that anyone can follow it, there will be no recovery when the next disaster strikes.

Lack of backup integrity. Many network administrators have a regularly scheduled backup to safeguard critical corporate data. But an oft-repeated worst practice is failing to ever validate the data or verify that you can restore it successfully and in a timely manner in the event of a disaster.

Self-defeating data storage practices. When it comes to backup and disaster recovery, offloading critical data from the network onto tapes or other removable media is a best practice. But it can become a worst practice in two ways:

One is by keeping the data on-site where it is sure to be destroyed along with the servers it backs up. In other words, what if the building burns down and takes your carefully stored data archives with it?

The second is by storing the removable media at another site that is either insecure itself or is where you cannot easily and quickly retrieve it when you need it.

No plan for backup power. Failing to plan for backup power in the event of a blackout or any other situation where power fails is a critical lapse. Lacking such a plan -- whether that means having sufficient capacity in uninterruptible power supply (UPS) systems to handle the load short term, or having a backup generator or other solution for longer outages -- incurs the possibility of going without power for an extended period of time.

No alternate facility. How quickly could you set up shop elsewhere? Not identifying an alternate site, and not making preparations for relocating to it in the event of a catastrophe, can mean not having a home for a while. Events like Hurricane Katrina demonstrate that it is important to not only have a backup facility in mind, but that you also make sure the backup facility is far enough away so as not to be affected by the same catastrophe.

No plan for communications. Not planning to have multiple methods of communicating with critical personnel is a worst practice for disaster recovery. If your plan calls for sending e-mail to a distribution list and the e-mail server crashes, there must be a Plan B in place. Similarly, if you plan to reach personnel by phone and the phone lines are down, you will need another way to reach them. As the 9/11 terrorist attacks demonstrated, cell phones can be unreliable too.

Not having the required computer and networking devices. Nothing will derail a backup and disaster recovery plan faster than not having the required computer and network equipment. Don't forget to identify standby equipment or another source that can provide or expedite the necessary servers, hard drives or other equipment if you need them.

No dry run. One of the most heinous practices is failing to test the plan. It may all look good on paper, but without a practice drill or dry run to validate each step of the plan, critical issues and contingencies may not get identified until it is too late.