Analysis: On the Road to Recovery
By Bob Lewis
Many of us have heard the expression, "Too soon old, too late smart." As it pertains to insuring that our computer installation, much less our company, can survive a major computer-related disaster, we should pay more attention to the truth of that adage.
Having spent the better part of the last 11 years trying to convince the powers that be that we should have some form of computer backup, it wasn't until several months ago that our firm's senior management became convinced that we NEEDED
a full Disaster Recovery (DR) plan. What convinced them is the realization that with the growth of an Internet site where our customers can purchase over 1,000 items online, and place over $4 billion in annual orders, there exists a potential for significant economic loss should a disaster befall our data center.
Several issues ago, I shared what I felt were some of the interesting aspects of being placed in charge of creating a DR plan for our firm, including some revelations as to the large differences in the cost of implementing such a plan. Now that we've been working for several months with our selected vendor, an update on our progress is in order.
I recently returned from a two-day class on the use of our vendor's DR software, as well as spending several days with their DR Consultant. What I've discovered is that the world of Disaster Recovery and Disaster Contingency planning is much more complex than I ever envisioned.
We're now in the first phase of implementing our DR plan, which consists of working with our vendor's DR consultant in the creation of a hardware and software inventory and selecting members of our Disaster Recovery team. In addition, we're in the process of determining if all critical data is being backed up, as well as calculating if current backup methods will allow restoration within acceptable time frames.
Once all these activities are completed, we'll proceed on to the next phase, which involves creating activation and notification plans, damage assessments, recovery procedures and finally conducting both a mock disaster exercise and an 18 hour "real" test at a regional hot site. It will take approximately five months to complete the entire plan, at which time we'll go into maintenance mode, which consists of using the DR software to keep our inventories of hardware, software and people up-to-date.
Although our firm has made a significant commitment, both in terms of manpower and monthly premiums, one only has to look at relevant disaster-related statistics to realize that NOT implementing a DR plan is NOT worth the possible risk.Average financial impact of one hour of data center down time:
(Source: Contingency Planning Research Inc., Livingston, N.J.)
- Telephone ticket sales $69,000
- Airline reservation centers $89,500
- Retail catalog sales centers $90,000Infomercial 800-number promotions $199,500
- Retail brokerage Firm $6.5 million
Along the way, I've uncovered some interesting disaster recovery facts to pass along.
- The computer backup industry started in the mid-1970s- with the creation of several regional hot sites.
- By the mid-1980s, there were over 100 providers of DR services.
- By 1989, a rapidly expanding client base generated over $240 million in annual subscription fees.
- Today, annual subscription revenues exceed $620 million.
- Three companies dominate the current DR market: SunGard, IBM, and Comdisco.
- The market is expanding at 17 percent per annum in revenues and 30 percent in subscriptions.
- Only 41 percent of U.S. companies have a DR plan in place.
A final piece of information from the U.S. Bureau of Labor you might find useful in convincing your management that a DR plan is worthwhile--93 percent of companies that suffer a significant data loss are out of business within five years. Don't be "too late smart."