Managing power consumption in data processing systems
Methods and systems for managing power consumption in data processing systems are disclosed. The power consumption in the data processing systems may be managed by obtaining power limits for the data processing systems. The power limits may be determined by computing the power limits from regularly updated parameters of the data processing systems. The data processing systems may include data processing systems that are housed in racks. From the regularly updated parameters, rack power limits may be obtained. Using the regular parameters and the rack power limits, data processing system power limits may be obtained for the data processing systems.
1 . A method for managing power consumption in data processing systems, the method comprising:
obtaining telemetry data for the data processing systems, the telemetry data comprising:
first data based on data processing system efficiencies of the data processing systems, and
second data based on power consumptions of the data processing systems;
obtaining data processing system efficiencies and the power consumptions for the data processing systems using the telemetry data;
obtaining, using the data processing system efficiencies and the power consumptions, normalized data processing system efficiencies and normalized power consumptions;
obtaining, using the normalized data processing system efficiencies and normalized power consumptions, weights for the data processing systems;
obtaining, using the weights for the data processing systems, power consumption limits for the data processing systems; and
updating operation of the data processing systems to prevent the data processing systems from exceeding the power consumption limits.
2 . The method of claim 1 , wherein the data processing system efficiencies indicate a level of workload performance of the data processing systems for a given level of power consumption by the data processing systems.
3 . The method of claim 2 , wherein a data processing system efficiency of the data processing system efficiencies is based, at least in part, on:
a hardware component loadout of a data processing system of the data processing systems;
configurations of the data processing system; and
historical performance of workloads by the data processing system.
4 . The method of claim 1 , wherein the power consumption limits impose maximum power usages by the data processing systems.
5 . The method of claim 1 , wherein obtaining the power consumption limits comprises:
obtaining, based on the weights for the data processing systems, rack weights for racks in which portions of the data processing systems are positioned;
obtaining, based on the rack weights and a total power budgeted for use by the data processing systems, rack power limits for the racks; and
obtaining, based on the rack power limits and the weights for the data processing systems, the power consumption limits for the data processing systems.
6 . The method of claim 5 , wherein the rack power limits impose maximum power usages by a rack of the racks, the rack housing a portion of the portions of the data processing systems.
7 . The method of claim 6 , wherein the rack weights prioritize power for allocation to racks that house the portions of the data processing systems.
8 . A non-transitory machine-readable medium having instructions stored therein, which when executed by a processor, cause the processor to perform operations for managing power consumption in data processing systems, the operation comprising:
obtaining telemetry data for the data processing systems, the telemetry data comprising:
first data based on data processing system efficiencies of the data processing systems, and
second data based on power consumptions of the data processing systems;
obtaining data processing system efficiencies and the power consumptions for the data processing systems using the telemetry data;
obtaining, using the data processing system efficiencies and the power consumptions, normalized data processing system efficiencies and normalized power consumptions;
obtaining, using the normalized data processing system efficiencies and normalized power consumptions, weights for the data processing systems;
obtaining, using the weights for the data processing systems, power consumption limits for the data processing systems; and
updating operation of the data processing systems to prevent the data processing systems from exceeding the power consumption limits.
9 . The non-transitory machine-readable medium of claim 8 , wherein the data processing system efficiencies indicate a level of workload performance of the data processing systems for a given level of power consumption by the data processing systems.
10 . The non-transitory machine-readable medium of claim 9 , wherein a data processing system efficiency of the data processing system efficiencies is based, at least in part, on:
a hardware component loadout of a data processing system of the data processing systems;
configurations of the data processing system; and
historical performance of workloads by the data processing system.
11 . The non-transitory machine-readable medium of claim 8 , wherein the power consumption limits impose maximum power usages by the data processing systems.
12 . The non-transitory machine-readable medium of claim 8 , wherein obtaining the power consumption limits comprises:
obtaining, based on the weights for the data processing systems, rack weights for racks in which portions of the data processing systems are positioned;
obtaining, based on the rack weights and a total power budgeted for use by the data processing systems, rack power limits for the racks; and
obtaining, based on the rack power limits and the weights for the data processing systems, the power consumption limits for the data processing systems.
13 . The non-transitory machine-readable medium of claim 12 , wherein the rack power limits impose maximum power usages by a rack of the racks, the rack housing a portion of the portions of the data processing systems.
14 . The non-transitory machine-readable medium of claim 13 , wherein the rack weights prioritize power for allocation to racks that house the portions of the data processing systems.
15 . A data processing system, comprising:
a processor; and
a memory coupled to the processor to store instructions, which when executed by the processor, cause the processor to perform operations for managing power consumption in data processing systems, the operations comprising:
obtaining telemetry data for the data processing systems, the telemetry data comprising:
first data based on data processing system efficiencies of the data processing systems, and
second data based on power consumptions of the data processing systems;
obtaining data processing system efficiencies and the power consumptions for the data processing systems using the telemetry data;
obtaining, using the data processing system efficiencies and the power consumptions, normalized data processing system efficiencies and normalized power consumptions;
obtaining, using the normalized data processing system efficiencies and normalized power consumptions, weights for the data processing systems;
obtaining, using the telemetry data, weights for the data processing systems;
obtaining, using the weights for the data processing systems, power consumption limits for the data processing systems; and
updating operation of the data processing systems to prevent the data processing systems from exceeding the power consumption limits.
16 . The data processing system of claim 15 , wherein the data processing system efficiencies indicate a level of workload performance of the data processing systems for a given level of power consumption by the data processing systems.
17 . The data processing system of claim 16 , wherein a data processing system efficiency of the data processing system efficiencies is based, at least in part, on:
a hardware component loadout of a data processing system of the data processing systems;
configurations of the data processing system; and
historical performance of workloads by the data processing system.
18 . The data processing system of claim 15 , wherein obtaining the power consumption limits comprises:
obtaining, based on the weights for the data processing systems, rack weights for racks in which portions of the data processing systems are positioned;
obtaining, based on the rack weights and a total power budgeted for use by the data processing systems, rack power limits for the racks; and
obtaining, based on the rack power limits and the weights for the data processing systems, the power consumption limits for the data processing systems.
19 . The data processing system of claim 18 , wherein the rack power limits impose maximum power usages by a rack of the racks, the rack housing a portion of the portions of the data processing systems.
20 . The data processing system of claim 19 , wherein the rack weights prioritize power for allocation to racks that house the portions of the data processing systems.