Analysis of IT equipment failures from high heat in data center
Data Center, Data Center administration, Data center cooling, Data Center in 2010, Server virtualization, Server Virtualization Implementation
Is there any analysis for the probability systems will fail as the temperature rises in a data center. Example, I have a CRAC unit fail, but the other is still running and cooling. I do not have enough tonage to keep the temperature at 74 degrees, so the room temperature will rise about a degree every minute. What is my risk of equipment failure as this temperature rises? I am looking for a way to measure the risk involved in not purchasing another CRAC unit as we work to virtualize our systems.
Software/Hardware used:
Mostly IBM equipment
Software/Hardware used:
Mostly IBM equipment
ASKED:
October 18, 2010 6:50 PM
UPDATED:
November 19, 2010 3:34 PM
RELATED QUESTIONS
Answer Wiki:
I have seen AS/400's and servers run in a 100 degree computer room like nothing was going on.
I have also seen(at IBM - Jackson, MS) a disk drive from a system where the building burnt down around it. They were able to pump the data from the drive.
That being said, get some fans on the equipment. Open the door if you have to. Try to keep it around 100 if not less.
To see all answers submitted to the Answer Wiki: View Answer History.
Let me revise. this question a little…my data center rises a degree every 30 mintues and I do apply fans in the room to help circulate the airflow out the only entrance to the data center…my question is more about a “study” that shows the probability systems will fail, with humidity and air flow being constant, when the temperature rises? I am looking for a way to measure the risk we are taking by not having an N+1 configuration in the data center.