Thermal monitoring and response apparatus and method for computer unit

• US 20060262474A1
• Filed: 05/23/2005
• Published: 11/23/2006
• Est. Priority Date: 05/23/2005
• Status: Active Grant

• Current Assignee: Network Appliance Incorporated (NetApp, Inc.)
• Original Assignee: Network Appliance Incorporated (NetApp, Inc.)
• Inventors: Ratty, Muthukumar, Chen, Chih-Yi, Xiong, Guohua
• Granted Patent: US 7,370,242 B2
• Time in Patent Office: Days
• US Class Current: 361/103
• CPC Class Codes: G06F 1/20 Cooling means
• CPC Class Codes: G06F 1/206 comprising thermal management

Abstract

Thermal conditions within a computer unit are monitored. A system-level temperature sensor responds to intake cooling air. A subsystem-level temperature sensor responds to the temperature of a subsystem-level component, preferably a motherboard or a backplane. When the temperature sensed by of the system-level temperature sensor falls outside of a critical limit, it is determined whether the temperature sensed by the subsystem-level temperature sensor generally correlates with the temperature sensed by the system-level temperature sensor. If there is no correlation, the operation of the computer unit is continued.

Claims

1. A method of monitoring thermal conditions within a computer unit which includes a system-level temperature sensor which responds to a temperature of intake cooling fluid for the computer unit and a subsystem-level temperature sensor which responds to a temperature condition of a subsystem-level component which is cooled by the cooling fluid within the computer unit, comprising:

determining whether the temperature sensed by of the system-level temperature sensor falls outside of a critical limit;

determining whether the temperature sensed by the subsystem-level temperature sensor generally correlates with the temperature sensed by the system-level temperature sensor when the temperature sensed by the system-level temperature sensor falls outside of the critical limit; and

shutting down operation of the computer unit only upon recognizing a correlation between the temperature sensed by the subsystem-level temperature sensor and the temperature sensed by the system-level temperature sensor when the temperature sensed by the system-level temperature sensor falls outside of the critical limit.

2. A method as defined in claim 1, further comprising:

shutting down operation of the computer unit only after the temperature sensed by the system-level temperature sensor remains outside of the critical limit for a predetermined time.

3. A method as defined in claim 2, further comprising:

continuing operation of the computer unit without shutting down operation of the computer unit if the temperature sensed by the system-level temperature sensor returns to within the critical limit within the predetermined time.

4. A method as defined in claim 1, further comprising:

continuing operation of the computer unit without shutting down operation of the computer unit if the temperature sensed by the system-level temperature sensor fails to correlate with the temperature sensed by the subsystem-level temperature sensor when the temperature sensed by the system-level temperature sensor falls outside of the critical limit.

5. A method as defined in claim 4, further comprising:

performing a diagnostic evaluation of the temperature sensors if the temperature sensed by the subsystem-level temperature sensor fails to correlate with the temperature sensed by the system-level temperature sensor when the temperature sensed by the system-level temperature sensor falls outside of the critical limit.

6. A method as defined in claim 1, further comprising:

obtaining a succession of values of the temperature sensed by the system-level temperature sensor when the temperature sensed by of the system-level temperature sensor falls outside of the critical limit;

determining whether a predetermined proportion of the succession of values of the temperature sensed by the system-level temperature sensor fall outside of the critical limit; and

evaluating whether to shut down operation of the computer unit upon the predetermined proportion of the succession of values of the temperature fall outside of the critical limit.

7. A method as defined in claim 6, further comprising:

continuing operation of the computer unit without shutting down operation of the computer unit if less than the predetermined proportion of the succession of values of the temperature sensed by the system-level temperature sensor fails to correlate with the temperature sensed by the subsystem-level temperature sensor when the temperature sensed by the system-level temperature sensor falls outside of the critical limit.

8. A method as defined in claim 7, wherein the intake cooling fluid is air and the system-level temperature sensor senses the temperature of the intake cooling air which flows through an enclosure of the computer unit, and the method further comprises:

determining whether the temperature of the intake cooling air is outside a predetermined warning limit but not outside of a predetermined critical limit which is beyond the predetermined warning limit; and

sending a warning message if the temperature of the intake cooling air is outside of the warning limit but not outside of the critical limit.

9. A method as defined in claim 8, further comprising:

obtaining a succession of values of the temperature of the intake cooling air; and

determining whether a predetermined proportion of the succession of values fall outside of the warning limit but not outside of the critical limit before sending the warning message.

10. A method as defined in claim 9, wherein the subsystem-level temperature sensor is a temperature sensor thermally associated with one of a motherboard or a backplane of the computer unit.

11. A method as defined in claim 1, wherein the computer unit includes a plurality of subsystem-level temperature sensors each of which is thermally associated with a different subsystem-level component of the computer system, and the method further comprises:

establishing a warning limit and a critical limit for the values of the temperatures sensed by each of the plurality of subsystem-level temperature sensors;

establishing each critical limit beyond each warning limit;

detecting a change of state upon the temperature sensed by one of any of the plurality of subsystem-level temperature sensors transitioning through one of the warning or critical limits;

obtaining a succession of values of the temperature sensed by the one subsystem-level temperature sensor for which the change of state has been detected;

determining whether a predetermined proportion of the succession of values obtained fall outside of the warning limit but not outside of the critical limit for that one subsystem-level temperature sensor; and

sending a warning message when the succession of values obtained fall outside of the warning limit but not outside of the critical limit.

12. A method as defined in claim 11, further comprising:

determining whether a predetermined proportion of the succession of values obtained fall outside of the critical limit for that one subsystem-level temperature sensor; and

sending a warning message if the succession of values obtained fall outside of the critical limit.

13. A method as defined in claim 12, further comprising:

continuing obtaining the succession of values of the temperature sensed by the one subsystem-level temperature sensor for which the warning message has been sent after sending the warning message; and

shutting down operation of the computer unit if the succession of values obtained after sending the warning message remains outside of the critical limit for a predetermined period of time.

14. A method as defined in claim 13, wherein the system-level temperature sensor comprises an intake cooling air temperature sensor for sensing the temperature of cooling air flowing through an enclosure of the computer unit, and the plurality of subsystem-level temperature sensors include a temperature sensor for sensing the temperature of one of a motherboard or a backplane within the computer unit.

15. A method as defined in claim 14, further including at least one component-level temperature sensor which is thermally associated with a different component-level component of the computer system, component-level component including a central processing unit within the enclosure of the computer unit, the method further comprising:

performing the aforesaid steps applicable to the subsystem-level temperature sensors with respect to the component-level component temperature sensors.

16. A method as defined in claim 15, wherein the plurality of subsystem temperature level sensors include at least one temperature sensor for sensing the temperature of at least one of a bus device connected to a communication bus within the computer unit or a remote LAN management device within the computer unit.

17. A method as defined in claim 1, wherein the system-level temperature sensor is a sensor for sensing the temperature of intake cooling air flowing into an enclosure of the computer unit for cooling components of the computer unit, and the subsystem-level temperature sensor is a temperature sensor thermally associated with one of either a motherboard or a backplane of the computer unit.

20. An apparatus as defined in claim 1, wherein the code executed by the processor:

obtains a succession of values of the temperature sensed by the system-level temperature sensor when the temperature sensed by the system-level temperature sensor falls outside of the critical limit threshold;

determines whether a predetermined proportion of the succession of values of the temperature sensed by the system-level temperature sensor fall outside of the critical limit threshold; and

commands shutting down operation of the computer unit upon the predetermined proportion of the succession of values obtained fall outside of the critical limit threshold.

21. An apparatus as defined in claim 20, wherein:

the cooling fluid is air;

the system-level temperature sensor is an intake cooling air temperature sensor for sensing the temperature of cooling air flowing through an enclosure of the computer unit; and

the computer unit includes a plurality of subsystem-level temperature sensors each of which is thermally associated with a different subsystem-level component of the computer unit; and

the code executed by the processor;

recognizes a change in state of the temperature sensed by one of the temperature sensors upon the temperature from that sensor transitioning through the warning limit threshold or the critical limit threshold;

detects a change of state of the temperature sensed by one of any of the plurality of subsystem-level temperature sensors;

obtains a succession of values of the temperature sensed by the one subsystem-level temperature sensor for which the change of state has been detected;

determines whether a predetermined proportion of the succession of values of the temperature sensed by the one subsystem-level temperature sensor fall outside the warning limit but not outside of the critical limit for that one subsystem-level temperature sensor; and

commands the delivery of a warning message when the succession of values of the temperature sensed by the one subsystem-level temperature sensor fall outside of the warning limit but not outside of the critical limit.

22. An apparatus as defined in claim 21, wherein the code executed by the processor:

determines whether a predetermined proportion of the succession of values of the temperature sensed by the one subsystem-level temperature sensor fall outside of the critical limit for that one subsystem-level temperature sensor; and

commands the delivery of a warning message if the succession of values of the temperature sensed by the one subsystem-level temperature sensor fall outside of the critical limit.

23. An apparatus as defined in claim 22, wherein the code executed by the processor:

commands shutting down operation of the computer unit after the temperature sensed by the one subsystem-level temperature sensor remains outside of the critical limit for a predetermined period of time.

24. An apparatus as defined in claim 23, wherein:

the computer unit further includes at least one component-level of temperature sensor which is thermally associated with a central processing unit (CPU) for sensing the temperature of a central processing unit within the enclosure of the computer unit.

25. An apparatus as defined in claim 24, wherein:

the plurality of subsystem temperature level sensors further include at least one of a temperature sensor for sensing the temperature of a bus device connected to a communication bus within the computer unit or a remote LAN management device within the computer unit.

18. An apparatus of monitoring thermal conditions within a computer unit which includes at least one of a motherboard or a backplane, comprising:

a system-level temperature sensor which supplies a temperature signal related to a temperature representative of intake cooling fluid for the computer unit;

a subsystem-level temperature sensor which supplies a temperature signal related to the temperature of one of either the motherboard or the backplane;

a memory storing values of warning limit thresholds and critical limit thresholds for the system-level temperature sensor and the subsystem-level temperature sensor, the critical limit threshold for each temperature sensor being beyond the warning limit threshold for that temperature sensor;

a processor connected to the memory and receptive of the temperature signals supplied by the system-level and subsystem-level temperature sensors, the processor executing code defining a process flow for monitoring thermal conditions of the computer unit in response to the temperature signals, and the code executed by the processor;

determines whether the temperature signal from the system-level temperature sensor falls outside of the critical limit threshold for the system-level temperature sensor;

determines whether the temperature signal from the system-level temperature sensor generally correlates with the temperature signal from the subsystem-level temperature sensor when the temperature signal from the system-level temperature sensor falls outside of the critical limit threshold; and

commands shutting down operation of the computer unit only upon recognizing a correlation between the temperature signal from the subsystem-level temperature sensor and the temperature signal from the system-level temperature sensor when the temperature signal from the system-level temperature sensor falls outside of the critical limit threshold for the system-level temperature sensor.

19. An apparatus as defined in claim 18, wherein the code executed by the processor:

commands shutting down operation of the computer unit only after the temperature sensed by the system-level temperature sensor remains outside of the critical limit threshold for a predetermined time.