IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0035320
(2005-01-14)
|
등록번호 |
US-7426453
(2008-09-16)
|
발명자
/ 주소 |
- Patel,Chandrakant D.
- Bash,Cullen E.
- Sharma,Ratnesh K.
|
출원인 / 주소 |
- Hewlett Packard Development Company, L.P.
|
인용정보 |
피인용 횟수 :
49 인용 특허 :
6 |
초록
▼
In a method of workload placement based upon capacity utilizations of a plurality of CRAC units, the provisioning of the plurality of CRAC units is determined and the zone of influence for each of the plurality of CRAC units is also determined based upon the provisioning of the plurality of CRAC uni
In a method of workload placement based upon capacity utilizations of a plurality of CRAC units, the provisioning of the plurality of CRAC units is determined and the zone of influence for each of the plurality of CRAC units is also determined based upon the provisioning of the plurality of CRAC units. It is determined whether a CRAC unit of the plurality of CRAC units is at least one of near failure or has failed. In addition, the workload is shifted from a zone of influence of at least one CRAC unit to a zone of influence of another of the plurality of CRAC units in response to a determination that a CRAC unit is at least one of near failure or has failed.
대표청구항
▼
What is claimed is: 1. A method of workload placement based upon capacity utilizations of a plurality of CRAC units, said method comprising: determining the provisioning of the plurality of CRAC units; determining the zone of influence for each of the plurality of CRAC units based upon the provisio
What is claimed is: 1. A method of workload placement based upon capacity utilizations of a plurality of CRAC units, said method comprising: determining the provisioning of the plurality of CRAC units; determining the zone of influence for each of the plurality of CRAC units based upon the provisioning of the plurality of CRAC units; determining whether a CRAC unit of the plurality of CRAC units is at least one of near failure or has failed; and shifting the workload from a zone of influence of at least one CRAC unit to a zone of influence of another of the plurality of CRAC units in response to a determination that a CRAC unit is at least one of near failure or has failed. 2. The method according to claim 1, further comprising: calculating actual loads on the plurality of CRAC units, calculating the difference between a sum of the actual loads and a sum of the rated capacities of the plurality of CRAC units; and wherein the step of shifting the workload comprises shifting the workload in response to the sum of the actual loads falling below the sum of the rated capacities. 3. The method according to claim 1, further comprising: determining whether the remaining plurality of CRAC units has sufficient capacity to redistribute the workload; and wherein the step of shifting the workload comprises shifting the workload in response to a determination that there is sufficient capacity in the remaining plurality of CRAC units. 4. The method according to claim 3, further comprising: calculating the actual available capacity in each of the remaining plurality of CRAC units in response to a determination that there is sufficient capacity; and wherein the step of shifting the workload comprises shifting the workload to the zones of influence of the remaining plurality of CRAC units based on the actual available capacities of the remaining plurality of CRAC units. 5. The method according to claim 4, further comprising: re-calculating actual loads on the remaining plurality of CRAC units; and determining whether the actual loads on the remaining plurality of CRAC units have reached the rated capacities of the remaining plurality of CRAC units. 6. The method according to claim 5, further comprising: determining whether additional workload is to be shifted in response to the actual loads an the remaining plurality of CRAC units falling below the rated capacities of the remaining plurality of CRAC units; and shifting the workload to the zones of influence of the remaining plurality of CRAC units based on the actual available capacities of the remaining plurality of CRAC units in response to a determination that additional workload is to be shifted. 7. The method according to claim 5, further comprising: determining the provisioning levels of the remaining plurality of CRAC units in response to the actual loads on the remaining plurality of CRAC units reaching the rated capacities of the remaining plurality of CRAC units; determining whether the provisioning levels of the CRAC units is safe; and scaling down power in servers of a zone of influence of a CRAC unit in response to a determination that the provisioning level of the CRAC unit is unsafe. 8. The method according to claim 7, wherein the servers have at least one processor, and wherein the step of scaling down power in servers comprises scaling down at least one or both of the performance and p-states of the at least one processor. 9. The method according to claim 1, further comprising projecting results of one or more workload shifting scenarios; generating at least one projected deployment map based upon projected results; selecting one of the at least one projected deployment maps; and wherein the step of shifting the workload comprises shifting the workload according to the selected one of the at least one projected deployment map. 10. The method according to claim 1, further comprising: generating a plurality of failure mitigation maps configured to model a plurality of workload distribution scenarios; and wherein the step of shifting the workload comprises shifting the workload according to one of the plurality of failure mitigation maps. 11. The method according to claim 1, wherein at least one of the plurality of CRAC units is located in a separate data center from the remaining plurality of CRAC units, and wherein the step of shifting the workload comprises shifting the workload to a zone of influence of the at least one of the CRAC units in the separate data center. 12. A method of providing sufficient capacity in a plurality of CRAC units to compensate for the failure of at least one of the plurality of CRAC units, said method comprising: determining the provisioning of the plurality of CRAC units; determining the zone of influence for each of the plurality of CRAC units based upon the provisioning of the plurality of CRAC units; calculating actual loads of the plurality of CRAC units; determining whether a sum of the actual loads equals or exceeds a predefined percentage of a sum of the rated capacities of the plurality of CRAC units; and scaling down power in servers in each zone of influence in response to the sum of the actual loads equaling or acceding the predefined percentage of the sum of the rated capacities, to thereby provide sufficient capacity in the plurality of CRAC units to compensate for the failure of at least one of the plurality of CRAC units. 13. The method according to claim 12, wherein the servers have at least one processor and wherein the step of scaling down power in servers comprises sealing down at least one or both of the performance and p-states of the at least one processor. 14. The method according to claim 12, wherein the step of scaling down power in the servers comprises scaling down power in the servers to a level cause the sum of the capacity utilizations to fall below the predefined percentage at the sum of the rated capacities. 15. A workload placement system comprising: a resource manager; a memory in communication with the resource manager, said memory comprising a module for calculating the actual loads of a plurality of CRAC units, said memory further comprising a module for determining placement of workload among a plurality of servers based upon the actual loads of the plurality of CRAC units; wherein the resource manager is configured to determine available capacities in the plurality of CRAC units in response to at least one of a failure and a potential failure in a CRAC unit of the plurality of CRAC units; and wherein the resource manager is further configured to shift the workload to at least one server in a zone of influence of at least one of the plurality of CRAC units having available capacity. 16. The workload placement system according to claim 15, wherein the resource manager is further configured to scale down power in one or more of the servers of a zone of influence of a CRAC unit. 17. The workload placement system according to claim 16, wherein the servers have at least one processor, and wherein the resource manager is configured to scale down at least one or both of the performance and p-states of the at least one processor. 18. The workload placement system according to claim 15, wherein the resource manager is further configured to generate at least one map depicting at least one projected deployment scenario, and wherein the resource manager is further configured to shift the workload based upon the at least one map. 19. The workload placement system according to claim 15, wherein at least one of the plurality of CRAC units is located in a separate data center from the remaining plurality of CRAC units, and wherein the resource manager is further configured to shift the workload to at least one server in a zone of influence of the at least one of the plurality of CRAC units located in the separate data center. 20. A data center comprising: a plurality of CRAC units,; means for determining the provisioning of the plurality of CRAC units; means for determining the zone of influence for each of the plurality of CRAC units; means for determining whether a CRAC unit of the plurality of CRAC units is at least one of near failure or has failed; and means for shifting the workload from a zone of influence of at least one CRAC unit to a zone of influence of another of the plurality of CRAC units based upon whether a CRAC unit has been determined to be at least one of near failure or has failed. 21. The data center according to claim 20, further comprising: means for scaling down at least one or both of the performance and p-states of at least one processor. 22. The data center according to claim 20, further comprising: means for generating at least one map depicting at least one projected deployment scenario, and wherein the means for shifting the workload is configured to shift the workload based upon the at least one map. 23. A computer readable storage medium on which is embedded one or more computer programs, said one or more computer programs implementing a method of workload placement based upon capacity utilizations of a plurality of CRAC units, said one or more computer programs comprising a set of instructions for: determining the provisioning of the plurality of CRAC units; determining the zone of influence for each of the plurality of CRAC units based upon the provisioning of the plurality of CRAC units; determining whether a CRAC unit of the plurality of CRAC units is at least one of near failure or has failed; and shifting the workload from a zone of influence of at least one CRAC unit to a zone of influence of another of the plurality of CRAC units in response to a determination that a CRAC unit is at least one of near failure or has failed. 24. The computer readable storage medium according to claim 23, said one or more computer programs further comprising a set of instructions for: calculating actual loads on the plurality of CRAC units, calculating the difference between a sum of the actual loads and a sum of the rated capacities of the plurality of CRAC units; and wherein the step of shifting the workload comprises shifting the workload in response to the sum of the actual loads falling below the sum of the rated capacities. 25. The computer readable storage medium according to claim 23, said one or more computer programs further comprising a set of instructions for: projecting results of one or more workload shifting scenarios; generating at least one projected deployment map based upon projected results; selecting one of the at least one projected deployment maps; and wherein the step of shifting the workload comprises shifting the workload according to the selected one of the at least one projected deployment map.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.