Data center load monitoring for utilizing an access power amount based on a projected peak power usage and a monitored power usage
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06F-001/32
G06F-001/26
출원번호
US-0134970
(2008-06-06)
등록번호
US-8949646
(2015-02-03)
발명자
/ 주소
Weber, Wolf-Dietrich
Fan, Xiaobo
Barroso, Luiz Andre
출원인 / 주소
Google Inc.
대리인 / 주소
Fish & Richardson P.C.
인용정보
피인용 횟수 :
4인용 특허 :
125
초록▼
A method of monitoring power load of a data center includes monitoring a power usage of a cluster of the data center, monitoring a power usage of a power distribution unit (PDU) of the data center, and monitoring a power usage of at least one of a rack of the data center or a computer within a rack
A method of monitoring power load of a data center includes monitoring a power usage of a cluster of the data center, monitoring a power usage of a power distribution unit (PDU) of the data center, and monitoring a power usage of at least one of a rack of the data center or a computer within a rack of the data center.
대표청구항▼
1. A method of managing a power load of a data center, comprising: monitoring a power usage of a cluster of the data center;monitoring a power usage of a power distribution unit (PDU) of the data center;monitoring a power usage of at least one of a rack of the data center or a computer within a the
1. A method of managing a power load of a data center, comprising: monitoring a power usage of a cluster of the data center;monitoring a power usage of a power distribution unit (PDU) of the data center;monitoring a power usage of at least one of a rack of the data center or a computer within a the rack of the data center;determining an excess power amount based on at least one of the monitored power usage of the cluster of the data center and a projected peak power usage of the cluster that exceeds a power capacity supplied to the cluster; or the monitored power usage of the PDU and a projected peak power usage of the PDU that exceeds a power capacity supplied to the PDU; andmanaging a manner in which computing jobs are executed at one or more of a plurality of computers in the data center to utilize the excess power amount and to provide for (1) compliance of the monitored power usage of the cluster with an available power capacity at the cluster, (2) compliance of the monitored power usage of the PDU with an available power capacity at the PDU, and (3) compliance of the monitored power usage of the at least one rack or the computer within the rack with an available capacity at the rack or the computer,wherein managing a manner in which computing jobs are executed at one or more of the plurality of computers in the data center comprises allocating the jobs among different computers within the plurality of computers in the data center and lowering an execution rate of a portion of the jobs having a lower priority than another portion of the jobs. 2. The method of claim 1, wherein monitoring the power usage of the cluster includes measuring power drawn through a power distribution system. 3. The method of claim 1, wherein monitoring the power usage of the cluster includes monitoring the utilization of at least some of a plurality of computers of the cluster and calculating power drawn by the plurality of computers from the utilization. 4. The method of claim 1, wherein monitoring the power usage of the PDU includes measuring power drawn through a power distribution system. 5. The method of claim 1, wherein monitoring the power usage of the PDU includes monitoring the utilization of at least some of a plurality of computers electrically coupled with the PDU and calculating power drawn by the plurality of computers from the utilization. 6. The method of claim 1, wherein monitoring a power usage of at least one of a rack of the data center or a computer within a rack of the data center comprises monitoring the power usage of the rack. 7. The method of claim 6, wherein monitoring the power usage of the rack includes measuring power drawn through a power distribution system by the rack. 8. The method of claim 6, wherein monitoring the power usage of the rack includes monitoring the utilization of at least some of a plurality of computers of the rack and calculating power drawn by the plurality of computers from the utilization. 9. The method of claim 1, wherein monitoring a power usage of at least one of a rack of the data center or a computer within a rack of the data center comprises monitoring the power usage of the computer within the rack. 10. The method of claim 9, wherein monitoring the power usage of the computer includes measuring power drawn by the computer. 11. The method of claim 9, wherein monitoring the power usage of the computer includes monitoring the utilization of the computer and calculating power drawn by the computer from the utilization. 12. The method of claim 1, further comprising monitoring power usage of subsets of a plurality of computers and calculating power drawn by the plurality of computers from the power drawn by the subsets. 13. The method of claim 12, wherein calculating power drawn by the plurality of computers includes summing the power drawn by the subsets. 14. The method of claim 1, further comprising monitoring a power usage of each computer in the data center. 15. The method of claim 1, further comprising monitoring a power usage of fewer than all computers in the data center. 16. The method of claim 1, further comprising measuring a power usage of each rack in the data center. 17. A method of managing a power load of a data center, comprising: determining a power usage for at least one cluster, for at least one power distribution unit, and for at least one rack arranged in the data center;determining an excess power amount based on at least one of: the power usage of the at least one cluster less than a peak projected power rating of the cluster,the power usage of the at least one power distribution unit less than a peak projected power rating of the power distribution unit, orthe power usage of the at least one rack less than a peak projected power rating of the at least one rack; andbased on the determination, managing a manner in which computing jobs are executed at one or more of a plurality of computers in the data center by allocating the computing jobs among different computers within the plurality of computers in the data center to utilize the excess power amount and by lowering an execution rate of a portion of the computing jobs having a lower priority than another portion of the computing jobs. 18. A method of managing a power load of a data center, comprising: storing data representing peak power usage of a plurality of computers in the data center, the peak power usage of the plurality of computers exceeding a power capacity supplied to the data center;collecting CPU utilization for at least a statistical sample of the plurality of computers;calculating a power load of the plurality of computers from the CPU utilization and the peak power usage; anddetermining an excess power amount based on a difference in the power load and the peak power usage;based on the calculated power load and the excess power amount, managing a manner in which computing jobs are executed at one or more of the plurality of computers in the data center by allocating the computing jobs among different computers within the plurality of computers in the data center and by lowering an execution rate of a portion of the computing jobs having a lower priority than another portion of the computing jobs. 19. A computer program product, tangibly embodied on a machine readable storage device, comprising instructions to cause a processor to: monitor a power usage of a cluster of a data center;monitor a power usage of a power distribution unit (PDU) of the data center;monitor a power usage of at least one of a rack of the data center or a computer within a rack of the data center;determine an excess power amount based on at least one of the monitored power usage of the cluster of the data center and a projected peak power usage of the cluster that exceeds a power capacity supplied to the cluster; or the monitored power usage of the PDU and a projected peak power usage of the PDU that exceeds a power capacity supplied to the PDU; andmanaging a manner in which computing jobs are executed at one or more of a plurality of computers in the data center to utilize the excess power amount and to provide for (1) compliance of the monitored power usage of the cluster with an available power capacity at the cluster, (2) compliance of the monitored power usage of the PDU with an available power capacity at the PDU, and (3) compliance of the monitored power usage of the at least one rack or the computer within the rack with an available capacity at the rack or the computer,wherein managing a manner in which computing jobs are executed at one or more of the plurality of computers in the data center comprises allocating the jobs among different computers within the plurality of computers in the data center and lowering an execution rate of a portion of the jobs having a lower priority than another portion of the jobs. 20. The product of claim 19, wherein allocating the jobs further comprises allocating the jobs to a subset of the racks having a lower power usage than remaining racks. 21. The method of claim 1, wherein managing a manner in which computing jobs are executed at one or more of the plurality of computers in the data center comprises scheduling the jobs to be executed by at least one of the plurality of computers in the data center. 22. The method of claim 21, wherein scheduling the jobs comprises: determining that a first subset of the jobs have a first priority;determining that a second subset of the jobs have a second priority, the second priority higher than the first priority; andscheduling the second subset of the jobs to begin execution by at least one of the plurality of computers in the data center prior to the first subset of the jobs. 23. The method of claim 1, wherein lowering the execution rate of the portion of the jobs having the lower priority than another portion of the jobs comprises providing an execution rate control.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (125)
Rasmussen, Neil; Simonelli, James; McNally, John; Brill, Kenneth G., Adjustable scalable rack power system and method.
Hillis,W. Daniel; Duttweiler,Mark; Salter,Kenneth D.; Yates,Randall A., Balanced chilled fluid cooling system for a data center in a shipping container.
Boros, Tibor, Distributed power management method for monitoring control/status signal of sub-modules to manage power of sub-modules by activating clock signal during operation of sub-modules.
Bahali,Sumanta K.; Bailey,Warren D.; Foster, Sr.,Jimmy G.; Sellman,Gregory D., Distributed system and method for managing power usage among server data processing systems.
Boss, Gregory J.; Doran, James R.; Hamilton, II, Rick A.; Sand, Anne R., Framework for distribution of computer workloads based on real-time energy costs.
Togawa, Atsushi, Method and apparatus for assigning tasks in an information processing system to optimize power consumption versus performance of the system.
Zwinger, Steven F.; Gross, Kenny C.; Dhanekula, Ramakrishna C., Method and apparatus for generating a dynamic power-flux map for a set of computer systems.
Brey,Thomas M.; Felter,Wesley M.; Lefurgy,Charles R.; Rajamani,Karthick; Rubio,Juan C.; Ware,Malcolm S., Method and system for providing performance estimations for a specified power budget.
Lefurgy, Charles R.; Govindan, Madhu Saravana Sibi, Method and system for real-time prediction of power usage for a change to another performance state.
Pomaranski,Ken Gary; Barr,Andrew Harvey, Method for determining number of dynamically temperature-adjusted power supply units needed to supply power according to measure operating temperature of power supply units.
Pomaranski,Ken Gary; Barr,Andrew Harvey, Method of providing dynamic power redundancy based on a difference of current power units and currently needed power units.
Hatasaki, Keisuke; Nakajima, Takao; Takamoto, Yoshifumi, Method, computing system, and computer program for reducing power consumption of a computing system by relocating jobs and deactivating idle servers.
Hatasaki, Keisuke; Nakajima, Takao; Takamoto, Yoshifumi, Method, computing system, and computer program for reducing power consumption of a computing system by relocating jobs and deactivating idle servers.
Baba,Takashige; Horimoto,Kazuhide, Multiple computer equipment and management method for determining number of AC-DC power modules to be operated by calculating power consumption based upon system information.
Baba, Takashige; Horimoto, Kazuhide, Multiple computer system equipped with a management module determining number of power supply modules to be operated based upon system information.
Felter,Wesley Michael; Keller, Jr.,Thomas Walter; Rajamani,Karthick; Rusu,Cosmin, Performance conserving method for reducing power consumption in a server system.
Bolan,Joesph E.; Brey,Thomas; Desai,Dhruv M.; Gruendler,Nickolas J.; Hughes,James E.; Klodnicki,Edward J.; Kochar,Sumeet; Shippy,Gary R., Power management using spare capacity of redundant power supply in blade environment.
Ali, Abid; Mitra, Asim, Scheduling of new job within a start time range based on calculated current load and predicted load value of the new job on media resources.
Paik, Nam Wook; Haines, Lance Palmer; Deokar, Vishwas Mohaniraj; Donalds, James H.; Manganese, Michael, System and method for allocating power to loads.
Loffink,John S.; Brundridge,Michael A.; Khatri,Mukund P.; Lambert,Timothy; Lopez,Pedro; Reddy,Jaydev, System and method for blade information handling system power inventory.
Hunt, Galen C.; Hydrie, Aamer; Welland, Robert V.; Tabbara, Bassam; Levi, Steven P.; Rehof, Jakob, System and method for designing a logical model of a distributed computer system and deploying physical resources according to the logical model.
Allison,Michael; Winick,Bradley D.; Zilavy,Daniel V.; Cross,Edward A.; Langley,Phillip David; Mankovich,James E., System and method for power management in a computer system having multiple power grids.
Layton, Jeffrey S.; Macnary, Reynolds R.; Moss, David L., System for determining servers power supply requirement by sampling power usage values thereof at a rate based upon the criticality of its availability.
Allen-Ware, Malcolm S.; Keller, Jr., Thomas W.; Lefurgy, Charles R.; Li, Yang; Rajamani, Karthick; Shanks, Samuel W.; Silva, Guillermo J.; Smith, Eddie L.; Yanes, James, Priority-aware power capping for hierarchical power distribution networks.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.