Power management of a computing system

1. A method for power management of a computing system, which comprises two or more physical servers for hosting virtual machines of a virtual system and one or more uninterruptible power supplies for supplying at least a subset of the physical servers with power, each of the one or more uninterruptible power supplies being connected to a phase of a multiple phase power supply, the method comprising:

receiving an action input for the computing system, which may impact the power consumption of the physical servers;

processing the received action input with a predictive model of power consumption of the physical servers regarding a battery autonomy of the one or more uninterruptible power supplies and a load balancing of several phases of the multiple phase power supply; and

optimizing utilization of the physical servers based on the result of the processing.

2. The method of claim 1 , comprising:

receiving measurements related to an operation of the physical servers;

using an artificial intelligence or machine learning algorithm for learning the power consumption of one or more individual parts of the computing system depending on actions and the measurements; and

generating and/or improving the predictive model of power consumption of the physical servers based on an output of the machine learning algorithm and the measurements.

3. The method of claim 2 , wherein the measurements related to the operation of the physical servers comprises at least one of the following:

total power consumption of the computing system;

temperature of an environment of the computing system;

virtual machines activity;

power consumption of physical servers;

processor activity of physical servers; and

mapping of virtual machines on the physical servers.

4. The method of claim 2 , wherein the machine learning algorithm receives a training data set based on the received measurements and a validation data set based on the received measurements and processes the training data set and the validation data set to generate the predictive model.

5. The method of claim 1 , wherein the optimizing of the utilization of the physical servers based on the result of the processing comprises:

receiving optimization constraints and optimization actions of the computing system;

determining one or more actions from the optimization actions for fulfilling the optimization constraints; and

using the determined one or more actions for the power management of the computing system.

6. The method of claim 5 , wherein the determining of one or more actions from the optimization actions for fulfilling the optimization constraints comprises determining a sequence of shutdown actions and/or shifting actions of virtual machines and/or physical servers depending on a remaining battery autonomy of the one or more uninterruptible power supplies and depending on the load balancing of the several phases of the multiple phase power supply.

7. A non-transitory computer-readable storage device storing software comprising instructions executable by a processor of a computing device which, upon such execution, cause the computing device to perform the method of claim 1 .

8. A system for power management of a computing system, which comprises two or more physical servers for hosting virtual machines of a virtual system and one or more uninterruptible power supplies for supplying at least a subset of the physical servers with power, each of the one or more uninterruptible power supplies being connected to a phase of a multiple phase power supply, the power management system comprising:

a predictive model of power consumption of the physical servers, the predictive model being provided to receive an action input for the computing system, which may impact the power consumption of the physical servers, and to process the received action input with regard to a battery autonomy of the one or more uninterruptible power supplies and a load balancing of several phases of the multiple phase power supply; and

an optimizer being provided for optimizing utilization of the physical servers based on the result of the processing by the predictive model.

9. The system of claim 8 , wherein the optimizer is provided to:

receive optimization constraints and optimization actions of the computing system;

determine one or more actions from the optimization actions for fulfilling the optimization constraints; and

use the determined one or more actions for the power management of the computing system.

10. The system of claim 9 , wherein the optimizer is provided to determine one or more actions from the optimization actions for fulfilling the optimization constraints by determining a sequence of shutdown actions of virtual machines and/or physical servers depending on a remaining battery autonomy of the one or more uninterruptible power supplies and depending on the load balancing of the several phases of the multiple phase power supply.