When Data Centers Go Wrong
Posted by Max Baun on Mon, Aug 25, 2008 @ 10:44 AM
As the IT Manager, your job is to effectively be in charge of all IT aspects while staying under budget as much as possible. You decide (or the decision is made for you) to cut back on expenses this year, trying to minimize the space you use for all of your servers and equipment by using one room or area.. In addition to cutting back on the physical space of your data center, you will get rid of the additional air conditioning unit that is not being used consistently to save on power expenses.
The next day you walk into the server room and immediately break out into a sweat. The tightly packed servers are generating so much heat that the temperature shot above 100 degrees overnight. Now the overworking cooling fans in the servers are damaging the equipment and the amount of IT help 
desk calls are skyrocketing. You decide to try and fit additional air conditioning into the room, but your servers are so jam packed that there is absolutely no more room to put anything.
Eventually, your equipment fails and your customers/users are filling up the phone lines with complaints. You’ve lost hundreds of thousands of dollars worth of equipment, not to mention the irreplaceable data that was stored on those machines. The servers that still work can’t be shut down because the company would then not be able to function at all. You call up the top corporations in the IT, HVAC, and power system industries and spend millions of dollars to re-create you’re the server room/ data center you once had.
The following links contains three situations where insufficient data center equipment leads to a horrible situation:
Pomona Valley Medical Center Example
It was consolidation time in the data center of Pomona Valley Medical Center. Due to jamming their servers together, the temperatures sky rocketed to above 100 degrees. One of the air conditioners was overworking and
eventually died.
After it was all said and done, they lost $40,000 worth of equipment. In the end, they had completely reform the data center and add 44 tons worth of air conditioning units.
Southern Nevada College Example
Southern Nevada College has roughly 40,000 students and 500 staff. With increase in students and staff, the need for more IT operations increased incredibly. Their closet-sized data center rooms started to experience more issues with temperature, humidity, wind, and power. All of these caused servers to fail and die. The best part is they were trying to cool the tightly squeezed equipment with their buildings central air conditioning system.
They ended up hiring Hewlett-Packard Co. and American Power Conversion Corp. to come and consolidate their equipment into one room and install virtualization software that they could monitor and control.
Norton Healthcare Example
The Norton Healthcare contains four hospitals, ten critical care locations, and fifty physician practice centers. Over the years, equipment, including more than 600 servers, was aimlessly added to their data center. Their equipment was drawing the maximum power that the center could handle which left no room 
for expansion. The walls and floor of their data center were hazardous due to its wooden construction and cut out ventilation holes. They also found old mainframe cabling under the floors which made everything worse.
The chief officers finally decided it was necessary to reform the data center and completely gutted the data rooms. They spent countless amount of money to completely rebuild the center, this time with the correct power distribution system, quality air conditioning systems, and correct security equipment.
What did the above situations have in common? They all thought they could save money and do it themselves. Their inadequate space and equipment caused their business stress that could have been avoided. What’s the lesson they all learned? Leave it up to the professionals.