Real-time inference of power efficiency metrics for a computer system
Some embodiments of the present invention provide a system that measures a power efficiency of a computer system. During operation, the system collects telemetry data from a set of sensors within the computer system. Next, the system determines a power consumption of the computer system from the telemetry data and determines a number of input/output operations per second (IOPS) for the computer system from the telemetry data. Finally, the system computes an IOPS per watt metric from the power consumption and the number of IOPS.
1. A method for measuring a power efficiency of a computer system, comprising:
collecting telemetry data from a set of sensors within the computer system;
determining a power consumption of the computer system from the telemetry data;
determining a number of input/output operations per second (IOPS) for the computer system from the telemetry data; and
computing an IOPS per watt metric from the power consumption and the number of IOPS by dividing the number of IOPS by a power consumption for storage devices for the computer system.
2. The method of claim 1 , wherein determining the power consumption of the computer system involves:
periodically polling current sensors and associated voltage sensors within the set of sensors to generate dynamic traces of currents and associated voltages for individual components within the computer system;
multiplying currents and associated voltages for the individual components within the computer system to produce dynamic traces of power consumption for the individual components; and
aggregating the dynamic traces of power consumption for the individual components to produce a dynamic trace of total power consumption for the computer system.
3. The method of claim 2 , wherein the current sensors and the associated voltage sensors are polled through a telemetry harness which measures sensor variables throughout the computer system.
4. The method of claim 3 , wherein a throughput script associated with the telemetry harness is used to determine the number of IOPS for the computer system.
5. The method of claim 1 , wherein determining the power consumption of the computer system involves applying an inferential model to the telemetry data to generate an inferential estimate of the power consumption from the telemetry data.
6. The method of claim 5 , wherein the inferential model is developed using a non-linear, non-parametric regression technique.
7. The method of claim 1 , wherein determining the number of IOPS for the computer system involves:
measuring a set of IOPS from a set of storage devices within the computer system; and
summing the IOPS measured from the set of storage devices.
8. A system for measuring a power efficiency of a computer system, comprising:
a telemetry harness configured to collect telemetry data from a set of sensors within the computer system; and
a power harness configured to:
determine a power consumption of the computer system from the telemetry data;
determine a number of input/output operations per second (IOPS) for the computer system from the telemetry data; and
compute an IOPS per watt metric from the power consumption and the number of IOPS by dividing the number of IOPS by a power consumption for storage devices for the computer system.
9. The system of claim 8 , wherein the power harness is configured to determine the power consumption of the system by:
obtaining dynamic traces of currents and associated voltages for individual components within the computer system from the telemetry data;
multiplying currents and associated voltages for the individual components within the computer system to produce dynamic traces of power consumption for the individual components; and
aggregating the dynamic traces of power consumption for the individual components to produce a dynamic trace of total power consumption for the computer system.
10. The system of claim 8 , wherein the power harness is configured to determine the power consumption of the system by:
applying an inferential model to the telemetry data to generate an inferential estimate of the power consumption from the telemetry data.
11. The system of claim 10 , wherein the inferential model is developed using a non-linear, non-parametric regression technique.
12. The system of claim 8 ,
wherein the telemetry harness comprises a throughput script configured to measure a set of IOPS from a set of storage devices within the computer system, and
wherein the power harness is configured to determine the number of IOPS for the computer system by summing the IOPS from the set of storage devices.
13. The system of claim 12 , wherein measuring the set of IOPS from the set of storage devices involves:
monitoring read and write operations from the storage devices; and
monitoring a number of bytes transferred during the read and write operations.
14. A computer-readable storage medium storing instructions that when executed by a computer cause the computer to perform a method for measuring a power efficiency of a computer system, the method comprising:
collecting telemetry data from a set of sensors within the computer system;
determining a power consumption of the computer system from the telemetry data;
determining a number of input/output operations per second (IOPS) for the computer system from the telemetry data; and
computing an IOPS per watt metric from the power consumption and the number of IOPS by dividing the number of IOPS by a power consumption for storage devices for the computer system.
15. The computer-readable storage medium of claim 14 , wherein determining the power consumption of the computer system involves:
periodically polling current sensors and associated voltage sensors within the set of sensors to generate dynamic traces of currents and associated voltages for individual components within the computer system;
multiplying currents and associated voltages for the individual components within the computer system to produce dynamic traces of power consumption for the individual components; and
aggregating the dynamic traces of power consumption for the individual components to produce a dynamic trace of total power consumption for the computer system.
16. The computer-readable storage medium of claim 15 , wherein the current sensors and the associated voltage sensors are polled through a telemetry harness which measures sensor variables throughout the computer system.
17. The computer-readable storage medium of claim 16 , wherein a throughput script associated with the telemetry harness is used to determine the number of IOPS for the computer system.
18. The computer-readable storage medium of claim 14 , wherein determining the power consumption of the computer system involves applying an inferential model to the telemetry data to generate an inferential estimate of the power consumption from the telemetry data.
19. The computer-readable storage medium of claim 18 , wherein the inferential model is developed using a non-linear, non-parametric regression technique.
20. The computer-readable storage medium of claim 14 , wherein determining the number of IOPS for the computer system involves:
measuring a set of IOPS from a set of storage devices within the computer system; and
summing the IOPS measured from the set of storage devices.