The Design Of Monitoring System And Analyze Local Hot Spot In IT-Room’s Thermal Environment

During the process of information construction, data center (IT-Room) plays animportant role in information management. A good IT-Room’s environment is thebase to guarantee the normal operation of equipment and prolong its lifetime.Traditional IT-Room’s monitoring method is to install some temperature sensorswhich are unable to cover all areas completely, so there are some problems about localhot spot. There are many factors which impact the area of local hot spot. Somemethods about finding local hot spot are on the basis of air distribution in IT-Roomand the inlet and outlet temperature of cabinet cross section. So the methods have acertain limit to find local hot spot.This article studies and analyzes the factors which impact the area of local hotspot by using Fluent Airpak to simulate the thermal environment of IT-Room. Amethod on the basis of curve-fitting is presented. The method takes some factors thatis power of machine inside cabinet, the temperature outlet minus inlet of cabinet crosssection, characteristics of air distribution, power of air-condition into consideration.On the basis of power of machine inside cabinet and the temperature outlet minusinlet of cabinet cross section, this method uses least squares method to fit BinomialCurve. It finds out local overheating cabinet through the rank of variance, whichcomes from machine inside cabinet whose fitted power and rated power.The local hot spot can be found and warned with limited sensor nodes, byapplying this method into the wireless sensor network monitoring system. The methodhas practical value and referential meaning in monitoring system, by analyzing andcomparing data about reality and simulation.