If the temperature exceeds the caution threshold for 60 seconds, a graceful server shutdown is
attempted.
•
Critical
—If temperatures are uncontrollable or rise quickly, the critical temperature threshold prevents
system failure by physically shutting down the server before the high temperature causes an electronic
component failure.
Monitoring policies differ depending on the server requirements. Policies usually include increasing fan
speeds to maximum cooling, logging temperature events in the IML, providing a visual indication of
events by using LED indicators, and starting a graceful shutdown of the operating system to avoid data
corruption.
Additional policies are implemented after an excessive temperature condition is corrected, including
returning the fan speed to normal, recording the event in the IML, turning off the LED indicators, and
canceling shutdowns in progress (if applicable).
Using the power and thermal features
165
