Determining a time for corrective action in a data center

Patent metadata

Field	Value
Publication number	US-9568923-B1
Application number	US-201514923647-A
Country	US
Kind code	B1
Filing date	Oct 27, 2015
Priority date	Oct 27, 2015
Publication date	Feb 14, 2017
Grant date	Feb 14, 2017

Abstract

Aspects include a method, system, and computer program product for determining a time to a threshold temperature of an electronic device in a data center. A method includes measuring parameters for an electronic device and the data center. A rate of change of temperature is determined for the device based on the parameters. The rate of change is compared to a rate of change threshold. It is determined that a cooling system is operating below a threshold when the rate of change is above the threshold. A first time is determined, where the first time is based the rate of change of temperature and a machine learning model. The first time and second time are compared, where the second time is a time to restore the cooling system to above the threshold. A signal is transmitted when the first time is less than the second time.

First claim

What is claimed is:

1. A computer implemented method for determining a time to a threshold temperature of an electronic device in a data center, the method comprising:
measuring with a plurality of sensors a plurality of operating parameters for a first electronic device and the data center;
determining a first rate of change of temperature for the first electronic device, the first electronic device being operated with a first workload, the first rate of change of temperature based at least in part on the plurality of operating parameters;
comparing the first rate of change to a rate of change threshold;
determining that a data center cooling system is operating below a performance threshold in response to the first rate of change being above the performance threshold;
determining a first time to a first temperature being above a device threshold, the first temperature associated with the first electronic device, the determination of the first time based at least in part on the first rate of change of temperature and a machine learning model;
comparing the first time to a second time, the second time being a time period for restoring the data center cooling system to above the performance threshold; and
transmitting a warning signal to an operator in response to the first time being less than the second time.

2. The method of claim 1 further comprising:
transmitting a cooling system signal to the operator in response to the first rate of change being above the rate of change threshold and the data center cooling system is operating at or above the performance threshold;
determining a cooling capacity level to reduce the first rate of change to below the rate of change threshold; and
changing an operation of the data center cooling system to generate at the cooling capacity level.

3. The method of claim 1 further comprising adding the first electronic device to a search domain in response to the first rate of change being below the rate of change threshold.

4. The method of claim 1 further comprising:
determining a second rate of change of temperature of a second electronic device, the second electronic device having a second workload and an associated capacity for workloads; and
comparing the second rate of change with the rate of change threshold.

5. The method of claim 4 further comprising:
determining a proposed rate of change of temperature for the second electronic device in response to the second rate of change being less than or equal to the rate of change threshold, the proposed rate of change of temperature being based on a transferring of a portion of the first workload from the first electronic device and added to the second workload;
comparing the proposed rate of change to the rate of change threshold;
comparing a sum of the portion of the first workload and the second workload to a capacity for workloads in response to the proposed rate of change being less than the rate of change threshold; and
moving the portion of the first workload to the second electronic device in response to the sum of the portion of the first workload and the second workload being less than or equal to the capacity for workloads.

6. The method of claim 5 further comprising:
removing the first electronic device from a search domain; and
removing the second electronic device from the search domain in response to the portion of the first workload and the second workload being greater than the capacity for workloads.

7. The method of claim 6 further comprising:
determining that the search domain includes additional electronic device; and
determining which of the additional electronic device has a lowest rate of change of temperature.

8. The method of claim 4 further comprising:
determining a third time to the first temperature being above the device threshold, the determination of the third time based at least in part on the first rate of change of temperature and the machine learning model; and
perform a mitigating step in response to determining the third time, the mitigating step selected from a group consisting of shutting down the electronic device, reducing a performance level of the electronic device and moving the first workload to a second data center.

9. A system for determining a time to a threshold temperature of an electronic device in a data center, the system comprising:
a memory having computer readable instructions; and
one or more processors for executing the computer readable instructions, the one or more processors being coupled to communicate with the plurality of sensors, the computer readable instructions comprising:
measuring with a plurality of sensors a plurality of operating parameters for a first electronic device and the data center;
determining a first rate of change of temperature for the first electronic device, the first electronic device being operated with a first workload;
comparing the first rate of change to a rate of change threshold, the first rate of change of temperature being based at least in part on the plurality of operating parameters;
determining that a data center cooling system is operating below a performance threshold in response to the first rate of change being above the rate of change threshold;
determining a first time to a first temperature being above a device threshold, the first temperature being associated with the first electronic device, the determination of the first time based at least in part on the first rate of change of temperature and a machine learning model;
comparing the first time to a second time, the second time being a time period for restoring the data center cooling system to above the performance threshold; and
transmitting a warning signal to an operator in response to the first time being less than the second time.

10. The system of claim 9 wherein the computer readable instructions further comprise:
transmitting a cooling system signal to the operator in response to the first rate of change is above the rate of change threshold and the data center cooling system is operating at or above the performance threshold;
determining a cooling capacity level to reduce the first rate of change to below the rate of change threshold; and
changing an operation of the data center cooling system to generate at the cooling capacity level.

11. The system of claim 9 wherein the computer readable instructions further comprise adding the first electronic device to a search domain the first rate of change is below the rate of change threshold.

12. The system of claim 9 wherein the computer readable instructions further comprise:
determining a second rate of change of temperature of a second electronic device, the second electronic device having a second workload and an associated capacity for workloads; and
comparing the second rate of change to the rate of change threshold.

13. The system of claim 12 wherein the computer readable instructions further comprise:
determining a proposed rate of change of temperature for the second electronic device in response to the second rate of change being less than or equal to the rate of change threshold, the proposed rate of change of temperature being based on a transferring of a portion of the first workload from the first electronic device and added to the second workload;
comparing the proposed rate of change to the rate of change threshold;
comparing a sum of the portion of the first workload and the second workload to a capacity for workloads in response to the proposed rate of change being less than the rate of change threshold; and
moving the portion of the first workload to the second electronic d

Assignees

• IBM

Inventors

• Demetriou Dustin W
• Schmidt Roger R
• Venkatesan Vidhya Shankar

Classifications

• G05D23/1902 Primary

  characterised by the use of a variable reference value · CPC title

• G05D23/1906

  using an analogue comparing device · CPC title

• G05D23/00

  Control of temperature · CPC title

• H05K7/20836

  Thermal management, e.g. server temperature control · CPC title

• H05K7/207

  Thermal management, e.g. cabinet temperature control · CPC title