IT Management Reference Guide

In these times of rising demands for high availability, coupled with heightened risks of unscheduled outages, the need for effective, affordable disaster recovery sites for data centers to ensure business continuity is becoming increasingly more important. A few months ago I had an assignment to evaluate a large number of alternative recovery sites for a financial services client. This four-part article presents as a case study the results of that evaluation which consisted of both a quantitative and qualitative analysis. In part one, I describe the first two of four quantitative comparisons I used: weather/climate issues and natural disaster issues.

This particular company has its primary data center located in Southern California in the city of Irvine. In addition to the Southern California region, we included the following 11 areas in this evaluation:

• Atlanta
• Dallas
• Denver
• Kansas City
• Jacksonville
• Las Vegas
• Northern California (San Francisco area)
• Phoenix
• Rolling Meadows (Chicago area)
• St. Louis
• White Plains (New York area)

We selected these areas because they were nation-wide which would allow us a large number of alternatives from which to select, and because they each contained major branch offices to provide some familiarity with the locales. I used the following four primary categories of characteristics and risk factors to conduct the analysis:

• weather/climate issues (such as snow, rain, humidity)
• natural disaster issues (such as earthquakes, floods, tornadoes)
• man-made disaster issues (such as electrical disturbances, terrorists)
• socio/economic issues (such as housing costs, wages and salaries)

Each of these categories contains several characteristics and risk factors that were applied to each of the 12 cities. Tables 1 and 2 list the results of our analysis for weather/climate issues and natural disaster issues. A variety of sources including websites, government agencies, recovery services companies, and industry specialists were used to compile this data.

For each factor within a category, the quantifiable entity associated with the factor is identified for each city. This quantity is then used to assign ranking points to each city in relation to every other city. For example, the first sub-category under weather and climate issues is the number of days in an average year that the temperature exceeded 90 degrees Fahrenheit. Phoenix had the highest number of days with 168, so it was rated 1 and designated as such with the parentheses. Las Vegas was second highest with 133, so it was rated 2, and so on for each city. A similar method was used for the other seven sub-categories. All of the ranking points were then summed. Cities with the lowest number of ranking points had the greatest level of issues within this category, and cities with the highest number of ranking points had the least level of issues.

As Table 1 shows, the three highest ranking areas in terms of having the least number of weather/climate issues were Southern California (80 points), Phoenix (75 points) and Las Vegas (71 points). A few points here are worth stressing. First, the category of weather/climate may or may not be of prime importance to you in selecting a recovery site. If not, do not use it and identify some other category that is more suitable for your particular situation. Similarly, you may want to use weather/climate but not with the sub-categories selected here. Determine other choices that make more sense for your environment. The important point here is that once you have selected your categories, use reliable information to evaluate the criteria and then rank the criteria accordingly. This technique of ranking and summing is used through the process for all of the categories selected.

Table 1 Weather/Climate Issues

Table 2 Disaster Issues During Past 10 Years

In part two I will present the results of my analysis for man-made and socio-economic issues.