Power profiling and auditing consumption systems and methods

1. A method for estimating power consumption in a network of computing devices, the method being performed by a processor of a primary server, the method comprising:

receiving operational information from a target server and a target server identifier, the operational information comprising parameter values aggregated over a period of time, a time stamp reflecting the end of the time period, and operational state of the target server during a time interval;

comparing elements of the received operational information with corresponding elements from a model of the target server that is selected in accordance with the target server identifier;

based on the comparison, identifying a set of elements from the model that most closely corresponds to the elements of the received operational information; and

estimating power consumption of the target server during the time interval based on the identified set of elements.

2. The method of claim 1 , wherein the operational information includes system-level data of the target server, including one or more of central processing unit (“CPU”) usage, disk activity and network activity.

3. The method of claim 2 , wherein the system-level data includes CPU usage measured as a percentage of time that a processor of the target server is active.

4. The method of claim 2 , wherein the system-level data includes disk activity measured as bytes transferred per second.

5. The method of claim 2 , wherein the system-level data includes network activity measured as bits transmitted or received every second.

6. The method of claim 2 , wherein the system-level data of the target server includes a cache hit/miss rate.

7. The method of claim 2 , wherein the system-level data of the target server further comprises processes that have been executed and their use of the CPU, disk activity and network activity.

8. The method of claim 1 , further comprising

determining whether an error exists in the estimated power consumption; and

when the error is present, recalibrating the model based on the received operational information.

9. The method of claim 1 , wherein the operational information is acquired by an agent installed in the target server.

10. The method of claim 9 , wherein the target server is a virtualized server.

11. The method of claim 10 , wherein the agent acquires at least some of the operational information from a hypervisor of a physical server that hosts the target server.

12. The method of claim 1 further comprising storing the estimated power consumption.

13. A system for estimating power consumption in a network, comprising:

a primary server configured to collect performance information from a plurality of target servers; and

a database configured to maintain benchmark performance data for at least one type of server and at least one configuration of each type of server, wherein the database is configured to organize the benchmark performance data for the each type and configuration of server for a plurality of operational conditions,

wherein the primary server estimates energy usage of the plurality of target servers using data extracted from the database based on the type of each target server and its operational conditions during a predetermined time period, and

wherein the benchmark performance data for each of the plurality of target servers is accessed using a target server identifier, and

wherein the primary server is configured to receive, from each target server, parameter values aggregated over a period of time, a time stamp reflecting the end of the time period, and operational state of the each target server during a time interval.

14. The system of claim 13 , wherein the plurality of target servers includes a virtualized server.

15. The system of claim 14 , wherein an agent is installed on each target server to acquire the performance information from the each target server and wherein at least some of the performance information collected from the virtualized server is collected from a hypervisor of a physical server that hosts the target server.

16. A non-transitory computer readable storage medium having one or more instructions that, when executed by a processor of a target server, cause the processor to:

acquire parameter values related to the target server upon expiration of a time interval, wherein causing the processor to acquire the parameter values includes causing the processor to make one or more system calls to the operating system of the target server;

aggregate the parameter values over the time interval into performance data;

transmit the performance data to a master server with a time stamp reflecting the end of the time interval and the operational state of the target server during the time interval, wherein the master server is configured to estimate power consumption of the target server by comparing the performance data to benchmark data maintained by the master server in a model of the target server; and

initiate a new time interval.

17. The storage medium of claim 16 , wherein the target server is a virtualized server.

18. The storage medium of claim 17 , wherein causing the processor to acquire parameter values includes causing the processor to obtain at least some of the performance data from a hypervisor of a physical server that hosts the target server.

19. The storage medium of claim 16 , wherein the performance data includes system-level data of the target server, including one or more of central processing unit (“CPU”) usage measured as a percentage of time that a processor of the target server is active, disk activity measured as bytes transferred per second and network activity measured as bits transmitted or received every second.

20. The storage medium of claim 19 , wherein the system-level data of the target server includes a cache hit/miss rate.