초록 ▼

Customers in a multi-tenant environment can obtain energy consumption information for a set of resources or other computing components used by those customers, including time-accurate accounting for various components of those resources utilized on behalf of the customer. A customer can also have th

Customers in a multi-tenant environment can obtain energy consumption information for a set of resources or other computing components used by those customers, including time-accurate accounting for various components of those resources utilized on behalf of the customer. A customer can also have the ability to specify how the resources are to be operated when used for the customer, in order to manage the amount of energy consumption. The accounting can be performed even when the resources are shared among multiple users or entities. Various hardware components or agents can be used to provide detailed energy consumption information for those components that is associated with a particular customer. The information can be used not only for accounting and monitoring purposes, but also to make dynamic adjustments based on various changes in usage, energy consumption, or other such factors.

대표청구항 ▼

1. A computer-implemented method for adjusting operational energy in a multi-tenant environment, comprising: determining a task to be performed for a first customer using at least one resource in the multi-tenant environment;determining a current operational state of the at least one resource for pr

1. A computer-implemented method for adjusting operational energy in a multi-tenant environment, comprising: determining a task to be performed for a first customer using at least one resource in the multi-tenant environment;determining a current operational state of the at least one resource for processing a task associated with a second customer, the current operational state including a value for at least one operational parameter of the at least one resource;storing information about the current operational state of the at least one resource associated with the second customer;setting the at least one resource to a requested operational state for the first customer when the requested operational state is different from the current operational state of the resource;determining an amount of energy consumption attributable to switching from the current operational state to the requested operational state;determining a first switching time for switching from the current operational state to the requested operational state for the first customer;determining a second switching time for switching from the current operational state to the requested operational state for the second customer;processing the task for the first customer using the at least one resource in the requested operational state; andattributing the amount of energy consumption of one or more components of the at least one resource attributable to switching from the current operational state to the requested operational state across at least one of the first customer or the second customer based at least in part on the first switching time and the second switching time. 2. The computer-implemented method of claim 1, wherein determining the amount of energy consumption includes reading at least one energy consumption counter for the at least one resource before and after switching from the current operational context to the requested operational context. 3. The computer-implemented method of claim 1, wherein storing the information about the current operational state of the at least one resource is performed in response to a machine instruction executed by the at least one resource. 4. The computer-implemented method of claim 3, wherein the machine instruction is further capable of causing the at least one resource to provide timing information for an amount of time in which a task is performed for at least one of the current operational state or the requested operational state. 5. The computer-implemented method of claim 1, wherein a context switch from the current operational state to the requested operational state is performed in response to an interrupt or machine instruction. 6. A computer-implemented method for adjusting energy consumption in a multi-tenant environment, comprising: determining a task to be performed for a customer using a resource in the multi-tenant environment;determining whether an operational state of the resource needs to be adjusted before performing the task for the customer, the operational state of the resource capable of having one or more state values when performing tasks for one or more other customers;adjusting the operational state of the resource when the operational state of the resource needs to be adjusted before performing the task for the customer;tracking a first amount of energy consumption of one or more components for the resource that is attributable to the adjusting of the operational state when the operational state of the resource needs to be adjusted, wherein the first amount of energy consumption comprises a first part having a corresponding first switching time and a second part having a second corresponding switching time;determining the first switching time for adjusting the operational state between state values for the customer;determining the second switching time for adjusting the operational state for the one or more other customers;tracking a second amount of energy consumption for the resource that is attributable to the performing of the task; andexposing energy consumption information for the task based at least in part upon the first part of the first amount of energy consumption for the resource that is attributable to the adjusting of the operational state when the operational state of the resource needs to be adjusted and the second amount of energy consumption for the resource that is attributable to the performance of the task for the customer. 7. The computer-implemented method of claim 6, wherein the requested operational state affects an amount of energy consumption of the resource. 8. The computer-implemented method of claim 6, wherein stored information about the current operational state is capable of being used by the resource to adjust the operational state of the resource back to the current operational state when another task requiring the current operational state is to be performed by the resource. 9. The computer-implemented method of claim 8, further comprising: tracking energy consumption by the resource attributable to the customer and any additional customers utilizing the resource. 10. The computer-implemented method of claim 9, wherein tracking the amount of energy consumption includes polling at least one energy consumption counter of the resource to determine amounts of energy consumed over a determined period of time. 11. The computer-implemented method of claim 10, wherein a value of the energy consumption counter is provided with information for the operational state to be stored before performing the task for the customer. 12. The computer-implemented method of claim 9, wherein determining the amount of energy consumption is based at least in part upon a number of instructions processed by the resource for the customer during a period of time. 13. The computer-implemented method of claim 12, wherein the number of instructions includes a first number of integer instructions and a second number of floating point instructions. 14. The computer-implemented method of claim 12, wherein there are at least two different classes of instructions, each of the at least two classes consuming energy in different amounts to complete execution. 15. The computer-implemented method of claim 14, wherein one or more different metrics are used to calculate energy consumption for each of the at least two different classes. 16. The computer-implemented method of claim 6, wherein the first amount of energy consumption for the resource that is attributable to the adjusting of the operational state includes energy consumption determined based upon at least one of an operating mode, operating state, operating frequency, or operating voltage of the resource during the adjusting of the operational state. 17. The computer-implemented method of claim 6, wherein the second amount of energy consumption for the resource includes energy consumption determined based upon at least one of an operating mode, operating state, operating frequency, or operating voltage of the resource during a period of time in which the resource is processing one or more instructions associated with the customer. 18. The computer-implemented method of claim 6, wherein a subsequent context switch is unable to be performed before at least a determined amount of energy is consumed that is a multiple of the first amount of energy consumption for the resource that is attributable to the adjusting of the operational state. 19. A computing system, comprising: at least one processor; andmemory storing instructions that, when executed by the at least one processor, cause the computing system to: receive, by an input/output interface, a request to perform a context switch for at least one resource, the context switch causing the at least one resource to go from a first operational state to a second operational state, each of the first and second operational states consuming energy at a different rate;perform, by the at least one processor, the context switch for the at least one resource;determine, by the at least one processor, an amount of energy consumption of one or more components for the at least one resource that is attributable to the context switch;determine, by the at least one processor, a first switching time for switching from the first operational state to the second operational state for at least a first customer;determine, by the at least one processor, a second switching time for switching from the first operational state to the second operational state for at least a second customer; andstore, by the at least one processor, the amount of energy consumption with respect to at least one of the first customer or the second customer based at least in part on the first switching time and the second switching time associated with the context switch. 20. The computing system of claim 19, wherein determining an amount of energy consumption for the at least one resource that is attributable to the context switch includes reading a first value from at least one energy consumption counter for the at least one resource before the context switch and a second value after the context switch and then calculating the amount of energy consumption attributable to the context switch based at least in part upon a sum of the differences between the first value and the second value for each energy consumption counter. 21. The computing system of claim 20, wherein a value of the at least one energy consumption counter is provided with information for the first operational state to be stored before performing the context switch. 22. The computing system of claim 19, wherein storing the amount of energy consumption is performed in response to a machine instruction executed by the at least one resource. 23. The computing system of claim 22, wherein the machine instruction is further capable of causing the at least one resource to provide timing information for an amount of time in which a task is performed for at least one of the first or second operational states. 24. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to: detect at least one condition prompting a context switch to be performed for at least one resource, the context switch causing the at least one resource to go from a first operational state to a second operational state, each of the first and second operational states consuming energy at a different rate;perform the context switch for the at least one resource;determine an amount of energy consumption of one or more components for the at least one resource that is attributable to the context switch;determining a first switching time for switching from the first operational state to the second operational state for at least a first customer;determining a second switching time for switching from the first operational state to the second operational state for at least a second customer; andstore the amount of energy consumption with respect to at least one of the first customer or the second customer based at least in part on the first switching time and the second switching time associated with the context switch. 25. The non-transitory computer-readable storage medium of claim 24, wherein the first operational state is associated with a task for the first customer and the second operational state is associated with a task for the second customer. 26. The non-transitory computer-readable storage medium of claim 25, wherein detecting the at least one condition includes detecting at least one of an amount of energy usage, a number of instructions processed, or an amount of time passing during operation in the first operational state. 27. The non-transitory computer-readable storage medium of claim 26, wherein the instructions when executed further cause the computing system to infer an average amount of power consumed by dividing the amount of energy usage by a corresponding amount of time. 28. The non-transitory computer-readable storage medium of claim 24, wherein determining an amount of energy consumption for the at least one resource that is attributable to the context switch includes reading a first value from at least one energy consumption counter for the at least one resource before the context switch and a second value after the context switch and then calculating the amount of energy consumption attributable to the context switch. 29. The non-transitory computer-readable storage medium of claim 24, wherein determining an amount of energy consumption for the at least one resource that is attributable to the context switch includes sampling a machine specific register (MSR) that records instantaneous power consumption. 30. The non-transitory computer-readable storage medium of claim 24, wherein storing the amount of energy consumption is performed in response to a machine instruction executed by the at least one resource, and wherein the machine instruction is further capable of causing the at least one resource to provide timing information for an amount of time in which a task is performed for at least one of the first or second operational states.