# US 2017/0255240 A1 · Energy efficient workload placement management based on observed server efficiency measurements > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2017/0255240 A1 - Document ID: US20170255240A1 - Status: Published application - Kind: A1 - Application number: 15/059,834 - Filed: 2016-03-03 - Published: 2017-09-07 - Earliest claimed: 2016-03-03 - Assignees: Lenovo Enterprise Solutions (Singapore) Pte. Ltd. - Applicants: Lenovo Enterprise Solutions (Singapore) Pte. Ltd. - Inventors: Jason A. Matteson; John W. Nicholson; Aparna Vallury; Scott Kelso - Web page: https://patentgrep.com/patent/US20170255240A1 ## CPC - G06F 1/26: Power supply means, e.g. regulation thereof (for memories G11C) - G06F 1/3209: Monitoring remote activity, e.g. over telephone lines or network connections - G06F 1/329: by task scheduling - G06F 1/3296: by lowering the supply or operating voltage - G06F 9/5094: {where the allocation takes into account power or heat criteria (power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206)} - Y02D 10/00: Energy efficient computing, e.g. low power processors, power management or thermal management - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - Y02D: CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE ## Abstract For each of a plurality of servers, a method includes obtaining current component power consumption values and calculating a current power consumption efficiency. The method further includes determining, for each of the plurality of servers, the current power consumption efficiency and an associated capacity utilization before and during performance of multiple instances of an identified workload. Then, for each server, the method determines a curve of power consumption efficiency as a function of capacity utilization that is representative of the performance of the plurality of instances of the identified workload. Embodiments of the method may then use the curve of power consumption efficiency curve in order to manage the power consumption efficiency of the plurality of servers. For example, the method may assign an additional workload to the server that is identified as having the greatest predicted power consumption efficiency. ## Claims ### Claim 1 A method, comprising: obtaining, for each of a plurality of servers, a plurality of current component power consumption values; calculating, for each of the plurality of servers, a current power consumption efficiency using the plurality of current component power consumption values; determining, for each of the plurality of servers, the current power consumption efficiency and a capacity utilization associated with the current power consumption efficiency before and during performance of a plurality of instances of an identified workload; and determining, for each of the plurality of servers, a curve of power consumption efficiency as a function of capacity utilization that is representative of the performance of the plurality of instances of the identified workload. ### Claim 2 (depends on 1) The method of claim 1, further comprising: obtaining an amount of capacity utilization necessary to perform an additional workload; determining, for each of the plurality of servers, a predicted power consumption efficiency by identifying a current capacity utilization, calculating a projected capacity utilization value by adding the amount of capacity utilization necessary to perform the additional workload to the current capacity utilization, and identifying the predicted power consumption efficiency that the curve associates with the projected capacity utilization value; identifying one of the plurality of servers that has the greatest predicted power consumption efficiency; and assigning the additional workload to the identified server. ### Claim 3 (depends on 2) The method of claim 2, wherein the workload is an additional workload that is not currently running on any of the plurality of servers. ### Claim 4 (depends on 2) The method of claim 2, wherein assigning the workload to the identified server including migrating the workload to the identified server from another one of the plurality of servers. ### Claim 5 (depends on 1) The method of claim 1, further comprising: obtaining an amount of capacity utilization necessary to perform an additional workload; determining, for each of the plurality of servers, a predicted power consumption efficiency by identifying a current capacity utilization, calculating a projected capacity utilization value by adding the amount of capacity utilization necessary to perform the additional workload to the current capacity utilization, and identifying the predicted power consumption efficiency that the curve associates with the projected capacity utilization value; determining, for each of the plurality of servers, any improvement in power consumption efficiency between the current power consumption efficiency and the predicted power consumption efficiency; identifying one of the plurality of servers that would experience the greatest improvement in power consumption efficiency; and assigning the workload to the identified server. ### Claim 6 (depends on 1) The method of claim 1, further comprising: measuring an actual change in power consumption efficiency as a result of performing the additional workload on the identified server; comparing the predicted power consumption efficiency change to the actual power consumption efficiency change; and modifying the curve of power consumption efficiency as a function of capacity utilization to better represent the actual change in power consumption efficiency as a result of performing the additional workload on the identified server; ### Claim 7 (depends on 1) The method of claim 1, wherein determining, for each of the plurality of servers, a curve of power consumption efficiency as a function of capacity utilization that is representative of the performance of the plurality of instances of the identified workload, includes performing a regression analysis of the power consumption efficiency and associated capacity utilization for the plurality of instances of the identified workload ### Claim 8 (depends on 1) The method of claim 1, wherein the capacity utilization is a processor utilization. ### Claim 9 (depends on 1) The method of claim 1, wherein the current power consumption efficiency is calculated as a ratio of total power input to the server and total power used to perform computing tasks. ### Claim 10 (depends on 1) The method of claim 1, wherein the current power consumption efficiency is an instantaneous current power consumption efficiency. ### Claim 11 (depends on 1) The method of claim 1, wherein the current power consumption efficiency is a moving average current power consumption efficiency. ### Claim 12 (depends on 1) The method of claim 1, further comprising: identifying the plurality of servers from among a larger network of servers, wherein the plurality of servers each has sufficient available capacity to perform the additional workload. ### Claim 13 (depends on 1) The method of claim 1, wherein the plurality of current component power consumption values include current power consumption values for two or more components selected from a fan, power supply, voltage regulator, processor, memory, data storage device, and input output adapter. ### Claim 14 (depends on 1) The method of claim 1, wherein a management server identifies the server having the greatest predicted power consumption efficiency. ### Claim 15 (depends on 1) The method of claim 1, further comprising: obtaining, for each of the plurality of servers, an ambient air temperature; and storing, for each of the plurality of servers, the ambient air temperature before and during performance of the identified workload in association with the stored current power consumption and the associated capacity utilization; wherein determining, for each of the plurality of servers, a curve of power consumption efficiency as a function of capacity utilization that is representative of the performance of the plurality of instances of the identified workload, includes determining, for each of the plurality of servers, a curve of power consumption efficiency as a function of capacity utilization that is representative of the performance of the plurality of instances of the identified workload at each of a plurality of the stored ambient air temperatures. ### Claim 16 (depends on 15) The method of claim 15, further comprising: measuring, for each of the plurality of servers, a current ambient air temperature, wherein identifying one of the plurality of servers that has the greatest predicted power consumption efficiency includes identifying one of the plurality of servers that has the greatest predicted power consumption efficiency at the current ambient air temperature. ### Claim 17 (depends on 1) The method of claim 1, wherein each server has one or more of a node manager, a baseboard management controller and an integrated management module for obtaining the plurality of current component power consumption values. ### Claim 18 (depends on 1) The method of claim 1, wherein obtaining an amount of capacity utilization necessary to perform the additional workload, includes storing a record of capacity utilization before and during performance of the workload each time the workload is performed. ### Claim 19 A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform a method comprising: obtaining, for each of a plurality of servers, a plurality of current component power consumption values; calculating, for each of the plurality of servers, a current power consumption efficiency using the plurality of current component power consumption values; determining, for each of the plurality of servers, the current power consumption efficiency and a capacity utilization associated with the current power consumption efficiency before and during performance of a plurality of instances of an identified workload; and determining, for each of the plurality of servers, a curve of power consumption efficiency as a function of capacity utilization that is representative of the performance of the plurality of instances of the identified workload. ### Claim 20 (depends on 19) The computer program product of claim 19, the method further comprising: obtaining an amount of capacity utilization necessary to perform an additional workload; determining, for each of the plurality of servers, a predicted power consumption efficiency by identifying a current capacity utilization, calculating a projected capacity utilization value by adding the amount of capacity utilization necessary to perform the additional workload to the current capacity utilization, and identifying the predicted power consumption efficiency that the curve associates with the projected capacity utilization value; identifying one of the plurality of servers that has the greatest predicted power consumption efficiency; and assigning the additional workload to the identified server.