Share what you know with millions of people
Focus is the best place to turn what you know into remarkable content
0
What is the most important factor in regional or national site selection for data centers?
Cost of power?
Economic incentives?
Ability to use economization?
Proximity to existing corporate locations?
I realize that different types of data centers will answer this question in a variety of ways.
Events
- Live now Dos and Don'ts of Small Business Marketing May 29 @ 11 am PT
- Lead Nurturing 202: The Next Generation May 31 @ 11 am PT
- The Tricks to Paid Media June 6 @ 11 am PT
- Display Advertising for Brand Awareness June 20 @ 11 am PT
Related Activity
10 Answers
Excellent question as well as the options that are listed for consideration: “Cost of power?”, “Economic incentives?”, “Ability to use economization?”, “Proximity to existing corporate locations?” Let’s see if I can address these and other important considerations.
If like most companies, your data center is the life blood of day-to-day operations. If your data center is down, what is your backup and disaster recovery plan? Because there are inherent risks from natural and manmade disasters, mitigation of those risks in the site selection of your data center should be priority one. All of the other considerations become moot if you select a site that is built on an active earthquake fault or prone to floods, hurricanes, tornados or wild fires. I’ve seen an insurance industry color coded map of the U.S. that rates locations based on their potential exposure to catastrophes. Needless to say the areas around the Gulf Coast, Florida, California and adjacent to the Mississippi are in red. So the first priority is picking a location that is least likely to be impacted by natural disaster, someplace like New Mexico or Arizona that doesn’t have blizzards, earthquakes, floods, tornados or hurricanes. An occasional monsoon only lasts minutes and rarely creates level of devastation of an F5 tornado or category five hurricane.
Once you decide on your primary location, pick a backup. This should be a redundant site if possible with full operational capacity that can support your operation in the event of failure or disaster.
If a fully redundant secondary site is too costly, you can create a redundant environment within your low-risk primary site. But at least you have made the choice to mitigate potential risks up front.
Now weight the options of available pool of resources to operate your center, the cost and availability of data communications, economic incentives and cost of power. In most cases you are going to find that some of these expenses will be offset by your reduced insurance premium.
Yes, all the “standard” site search criteria are important but...."Bandwidth" is tops.
Because most site search consultants and brokers don't have a clue what bandwidth is, how it figures into total cost, how performance requirements differ between end users, where we are headed and how mobile computing is changing the way we work, etc. almost everyone has written this off. Don’t mistake this for the rather silly questions “Where is the fiber?” or “Do you have more then one fiber provider?
A data center is a long term investment and Network Planning needs to think, five, ten, fifteen years ahead.
I agree with Bradford and will state it again briefly. The most important factor in site selection should be survivability. You must have a location that geographically mitigates the risk of earthquake, flood, severe weather, and many other natural disasters that can be identified on a regional basis
There are many, but I see electric grid options and carrier diversity as paramount to location selection. Some other important points have been listed already.
Recently went through this exercise on a global basis. We had a few key criteria for selecting candidates: natural disasters, stability of local government and economy, fiber density, network hub locations, tier 3 or higher, available space, location relative to our facilities. In the US, that rules out CA, FL and a few other locations. We looked at Chicago, Phoenix, Las Vegas, Colorado and NJ. In Asia, Singapore was one of the better options. In Europe, Germany was our preferred choice. In the UK, the centers had limited space and local resources are pricey there.
All great points! I really liked Bradford's answer! he hit all the most important points for when selecting a data center site. You do have to make sure that you can survive or be least closest to any potential disaster. Great posts all in all.
When we select facilities we tend to lean towards the major telco collocation facilities in the area and build our infrastructure within close proximity to the rest of the providers, this way we are located in a very technically busy environment and have much knowledge around us. These locations are typically in high demand and commonly more costly; however we purchase large cages or suites, power, bandwidth and insert our technical staff to support our equipment as well as our customer’s cabinets & equipment.
Many smaller data centers are limited to one or a few providers and are typically isolated due to lower lease or building acquisition costs.
You need to ask, what is the level of support you require & how much space do you need? This will help determine a more ideal partner to work with.
Survivability is everyone’s number one criteria. But it is also the most commonly ignored requirement.
If the facility is priced right then the fact that the structure has less than the required wind rating, or the building sits just off of an airport runway, or railroad tracks, or oil pipeline, or is in a downtown area and cannot be protected, has public parking in the same building, etc. gets rationalized away in all of 30 seconds.
There are many examples all over the country of facilities that break one or more rules of survivability.
I’m not suggesting that survivability should not be one of the highest criteria but in reality it seldom is, which makes me ask “Why is survivability only a token requirement for most organizations?"
Ken, The problem you are highlighting is, I think, due to the fact most disaster recovery/business continuity site selections are made with natural disasters in mind such as floods and tornados or utility failures such as carrier or electric company. I don't think total destruction or nuclear holocaust are part of the selection process. Maybe they should they be. It depends on each company's circumstances and goals for disaster avoidance. The number of single points of failure that one can address is directly relational to the amount of money one has to spend. Get as much survivability as you possibly can for the best price available.
every nations should have datawarehouse, and displayed on applications like dashboard or GIS
Answer This Question