Dynamic server-level thermal control based on component temperatures
An apparatus for monitoring and maintaining thermal exhaust temperatures for monitored servers may include a processor in communication with an exhaust temperature control system. The processor may implement various control loops to monitor inlet and exhaust temperatures for one or more servers, and the processor may monitor temperatures of various hardware components within those one or more servers. When the processor determines that a hardware component for a given server is overheating, the processor may take steps to decrease the temperature of the overheating hardware component for the given server. To determine whether a hardware component is overheating, the processor may leverage the hardware component's operating temperature margin.
1. A method for controlling an exhaust temperature by establishing temperature difference values, the method, implemented by a processor, comprising:
monitoring, with the processor in communication with an exhaust temperature control system, an inlet temperature value associated with a plurality of components;
determining a temperature difference value based on the inlet temperature value and a stored association between the inlet temperature value and a stored temperature difference value;
determining an exhaust temperature value based on the inlet temperature value and the temperature difference value;
transmitting the exhaust temperature value to the exhaust temperature control system;
determining a temperature of each component of the plurality of components;
determining a plurality of component temperature margin values by determining, for each component of the plurality of components, a component temperature margin value based on the difference between a temperature of the component and a threshold operating temperature associated with the component;
establishing the temperature difference value based on the lowest component temperature margin value of the plurality of component temperature margin values; and
transmitting the established temperature difference value to the exhaust temperature control system to control the exhaust temperature.
2. The method of claim 1 , wherein the stored association and the stored temperature difference values are being stored in a memory prior to said monitoring.
3. The method of claim 1 , wherein the inlet temperature value comprises an inlet temperature value derived from an inlet temperature source.
4. The method of claim 1 , wherein the inlet temperature value is selected based on a geographic location.
5. The method of claim 1 , wherein the inlet temperature value is determined by averaging a plurality of temperature measurements.
6. The method of claim 1 , wherein determining the temperature difference value comprises querying a temperature difference database with the inlet temperature value.
7. The method of claim 1 , further comprising increasing the temperature difference value by a predetermined amount based on a comparison of an average value of the determined component temperature margin values with a second component margin threshold value.
8. The method of claim 6 , further comprising determining the temperature difference value through interpolation when the temperature difference database does not include the inlet temperature value.
9. The method of claim 3 , wherein the inlet temperature source is a temperature probe that monitors an air inlet.
10. The method of claim 3 , further comprising polling the inlet temperature source at predetermined time intervals to update the inlet temperature value.
11. The method of claim 1 , further comprising:
associating each component temperature margin value with at least one of a plurality of components; and
associating each component with at least one of a plurality of zones.
12. The method of claim 11 further comprising updating a zone database when there is a change to a given zone, wherein the change is selected from a group of changes consisting of adding, removing, and replacing a component.
13. The method of claim 11 , wherein each of the plurality of zones is associated with one or more cooling devices.
14. The method of claim 1 , further comprising operating the one of the plurality of components at increasing temperatures until the one of the plurality of components exhibits a decrease in performance or an instability.