Dynamic compute composition
US-2024311210-A1 · Sep 19, 2024 · US
US9804897B2 · US · B2
| Field | Value |
|---|---|
| Publication number | US-9804897-B2 |
| Application number | US-201314383535-A |
| Country | US |
| Kind code | B2 |
| Filing date | Mar 6, 2013 |
| Priority date | Mar 6, 2012 |
| Publication date | Oct 31, 2017 |
| Grant date | Oct 31, 2017 |
A practical reading order for non-experts. Skip the full description unless you need deep technical detail.
What the patent document calls the invention.
A short plain-language summary of the technical disclosure.
Who owns or filed the patent and who is credited as inventor.
Filing, priority, publication, and grant dates set the timeline.
The legal scope of protection — read this for what is actually claimed.
Technology tags used to group this patent with similar filings.
Prior art links and similar publications in this corpus.
Official abstract text for this publication.
Provided are a power management method and apparatus that minimize power consumption in a virtualization system using different operating systems. The power management method operates on a virtualization layer, and may include: monitoring usage states of real processors actually performing tasks and usage states of virtual processors mapped with the real processors and executing operating systems or applications; predicting workload of a virtual processor; computing the utilization level of a real processor enough for supporting the predicted workload; and reconfiguring the mapping between real processors and virtual processors according to the computation result. Hence, different power management schemes of multiple operating systems or applications are managed at the virtualization layer in an integrated manner, preventing a conflict caused by application of different power management schemes. Power consumption in the overall system can be minimized through prediction of resource usage.
Opening claim text (preview).
The invention claimed is: 1. A method for power management on a virtualization layer in a virtualization system using different operating systems, the method comprising: monitoring usage states of real processors performing tasks and usage states of virtual processors mapped with the real processors, the virtual processors executing operating systems or applications; determining workload of a virtual processor, determining a utilization level of a real processor for supporting the determined workload; and reconfiguring the mapping between real processors and virtual processors according to the utilization level of the real processors; wherein determining the workload of the virtual processor comprises: determining errors of a weighting algorithm and a simple proportional integral differential (PID) algorithm; identifying whether a selected algorithm is present; and selecting, when the selected algorithm is not present, another algorithm producing a smaller error. 2. The method of claim 1 , wherein determining the workload of the virtual processor further comprises: repeatedly identifying, when the selected algorithm is present, a given number of times whether an error of the selected algorithm is larger than that of a non-selected algorithm; and replacing, when for the given number of times in a row the error of the selected algorithm is larger than that of the non-selected algorithm, the selected algorithm. 3. The method of claim 2 , wherein determining the workload of the virtual processor further comprises maintaining, when for the given number of times in a row the prediction error of the selected algorithm is smaller than or equal to that of the non-selected algorithm, the selected algorithm. 4. The method of claim 1 , wherein reconfiguring the mapping between real processors and virtual processors comprises: determining a possibility of handling the determined workload by use of a currently configured mapping; and reconfiguring, upon determining that the handling possibility is low, the mapping by allocating an additional real processor. 5. The method of claim 4 , wherein reconfiguring the mapping between real processors and virtual processors further comprises: determining, upon determining that the handling possibility is high, necessity of increasing the utilization level of the real processor with respect to a current utilization level; and decreasing, when not necessary to raise the utilization level, the utilization level of the real processor according to the determined workload. 6. The method of claim 5 , wherein reconfiguring the mapping between real processors and virtual processors further comprises increasing, when necessary to increase the utilization level, the utilization level of the real processor in accordance with the determined workload. 7. The method of claim 1 , wherein reconfiguring the mapping between real processors and virtual processors comprises adding margins to the determined workload so as to prevent an abrupt change in the workload. 8. An apparatus for power management in a virtualization system using different operating systems, the apparatus comprising: real processors configured to perform tasks; virtual processors mapped with the real processors and configured to execute operating systems or applications; and a controller configured to monitor usage states of the real processors and the virtual processors, on a virtualization layer, determine workload of a virtual processor, determine a utilization level of a real processor for supporting the determined workload, reconfigure the mapping between the real processors and virtual processors according to the utilization level of the real processors, determine errors of a weighting algorithm and a simple proportional integral differential (PID) algorithm, identify whether a selected algorithm is present, and select, when a selected algorithm is not present, another algorithm producing a smaller error. 9. The apparatus of claim 8 , wherein, when a selected algorithm is present, the controller is configured to repeatedly identify a given number of times whether an error of the selected algorithm is larger than that of a non-selected algorithm, and replace, when for the given number of times in a row the error of the selected algorithm is larger than that of the non-selected algorithm, the selected algorithm. 10. The apparatus of claim 9 , wherein, when for the given number of times in a row the error of the selected algorithm is smaller than or equal to that of the non-selected algorithm, the controller is configured to maintain the selected algorithm. 11. The apparatus of claim 8 , wherein, for mapping reconfiguration, the controller is configured to check a possibility of handling the determined workload by use of a currently configured mapping, and reconfigure, upon determining that the handling possibility is low, the mapping by allocating an additional real processor. 12. The apparatus of claim 11 , wherein, upon determining that the handling possibility is high, the controller is configured to determine necessity of increasing the utilization level of the real processor with respect to a current utilization level, and decrease, when not necessary to increase the utilization level, the utilization level of the real processor according to the determined workload. 13. The apparatus of claim 12 , wherein, when necessary to increase the utilization level, the controller is configured to increase the utilization level of the real processor in accordance with the determined workload. 14. The apparatus of claim 8 , wherein the controller is configured to reconfigure the mapping by adding margins to the determined workload so as to prevent an abrupt change in the workload. 15. A method for power management on a virtualization layer in a virtualization system using different operating systems, the method comprising: monitoring usage states of one or more virtual processors mapped with real processors, the virtual processors executing operating systems or applications; determining errors of a weighting algorithm and a simple proportional integral differential (PID) algorithm; identifying whether a selected algorithm is present; selecting, when the selected algorithm is not present, another algorithm producing a smaller error; determining workload of the one or more virtual processors based on the selected algorithm; identifying an operation state of hardware equipment based on the workload of the one or more virtual processors; and changing, when the hardware equipment is concurrently used by multiple virtual processors, the operation state of the hardware equipment based on usage state of the hardware equipment. 16. The method of claim 15 , further comprising changing, when the hardware equipment is not concurrently used by multiple virtual processors, the operation state of the hardware equipment according to a received operation state change command. 17. An apparatus for power management in a virtualization system using different operating systems, the apparatus comprising: one or more virtual processors mapped with real processor configured to perform tasks, the virtual processors configured to execute operating systems or applications; a hardware equipment configured to perform an operation indicated by a command from a virtual processor; and a controller configured to: monitor usage states of the virtual processors, on a virtualization layer; determine errors of a weighting algorithm and a simple proportional integral differential (PID) algorithm; identify whethe
Logical partitioning of resources; Management or configuration of virtualized resources (specific details on emulation or internal functioning of virtual machines G06F9/455) · CPC title
Hypervisor-specific management and integration aspects · CPC title
Cross-Sectional Technologies · mapped topic
Workload prediction · CPC title
Distribution of virtual machine instances; Migration and load balancing · CPC title
Related publications grouped by family.
Answers are generated from the same data shown on this page.