OPTIMIZING A PREDICTION OF RESOURCE USAGE OF MULTIPLE APPLICATIONS IN A VIRTUAL ENVIRONMENT

• US 20100083248A1
• Filed: 09/30/2008
• Published: 04/01/2010
• Est. Priority Date: 09/30/2008
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Cherkasova, Ludmila, Wood, Timothy W.
• Granted Patent: US 8,180,604 B2
• Time in Patent Office: Days
• US Class Current: 718/1
• CPC Class Codes: G06F 11/3428 Benchmarking
• CPC Class Codes: G06F 11/3442 for planning or managing the needed capacity
• CPC Class Codes: G06F 11/3447 Performance evaluation by modeling
• CPC Class Codes: G06F 11/3466 Performance evaluation by tracing or monitoring
• CPC Class Codes: G06F 2009/45591 Monitoring or debugging support
• CPC Class Codes: G06F 2201/815 Virtual
• CPC Class Codes: G06F 2201/865 Monitoring of software
• CPC Class Codes: G06F 2209/5019 Workload prediction
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/5077 Logical partitioning of resources; Management or configuration of virtualized resources specific details on emulation or internal functioning of virtual machines G06F9/455

Abstract

Described herein is a method for optimizing a prediction of resource usage of multiple applications running in a virtual environment, comprising: providing a predetermined set of benchmarks; executing the predetermined set of benchmarks in a native hardware system in which the application natively resides; executing the predetermined set of benchmarks in the virtual environment; collecting first traces of first resource utilization metrics in the native hardware system based on the execution of the predetermined set of benchmarks in the native hardware system; collecting second traces of second 10 resource utilization metrics in the virtual environment based on the execution of the predetermined set of benchmarks in the virtual environment; generating a first prediction model that maps a first selected set of the first traces of a selected one of the first resource utilization metrics to the second traces of resource utilization metrics; generating a second prediction model that maps a second different selected set of the first traces of the selected first resource utilization metric to the second traces of resource utilization metrics; collecting first application traces of resource utilization metrics in the native hardware system based on an execution of a first application in the native hardware system; collecting second application traces of resource utilization metrics in the native hardware system based on an execution of a second application in the native hardware system; aggregating the first application traces of the first application and the second application traces of the second application into combined application traces of resource utilization metrics; and predicting a combined resource usage of the first and second applications running in the virtual environment by applying the first and second prediction models to the combined application traces of resource utilization metrics.

Claims

1. A method for optimizing a prediction of resource usage of multiple applications running in a virtual environment, comprising:

providing a predetermined set of benchmarks;

executing the predetermined set of benchmarks in a native hardware system in which the application natively resides;

executing the predetermined set of benchmarks in the virtual environment;

collecting first traces of first resource utilization metrics in the native hardware system based on the execution of the predetermined set of benchmarks in the native hardware system;

collecting second traces of second resource utilization metrics in the virtual environment based on the execution of the predetermined set of benchmarks in the virtual environment;

generating a first prediction model that maps a first selected set of the first traces of a selected one of the first resource utilization metrics to the second traces of resource utilization metrics;

generating a second prediction model that maps a second selected set of the first traces of the selected first resource utilization metric to the second traces of resource utilization metrics;

collecting first application traces of resource utilization metrics in the native hardware system based on an execution of a first application in the native hardware system;

collecting second application traces of resource utilization metrics in the native hardware system based on an execution of a second application in the native hardware system;

aggregating the first application traces of the first application and the second application traces of the second application into combined application traces of resource utilization metrics; and

predicting a combined resource usage of the first and second applications running in the virtual environment by applying the first and second prediction models to the combined application traces of resource utilization metrics.

2. The method of claim 1, wherein applying the first and second prediction models comprises:

applying each of the first and second prediction models to the combined traces of resource utilization metrics based on an input/out rate of a selected one of the resource utilization metrics.

3. The method of claim 1, further comprising:

generating a third prediction model that maps the first and second selected sets of the first traces of the selected first metric to the second traces of resource utilization metrics; and

predicting a combined resource usage of the first and second applications running in the virtual environment by applying the third prediction model to each of the collected first and second traces of resource utilization metrics.

4. The method of claim 3, wherein predicting the combined resource usage by applying the third prediction model comprises:

applying the third prediction model to the collected first application traces of resource utilization metrics to estimate a first resource usage of the first application running in the virtual environment;

applying the third prediction model to the collected second application traces of resource utilization metrics to estimate a second resource usage of the second application running in the virtual environment; and

predicting the combined resource usage of the first and second applications by aggregating the first and second resource usage.

5. The method of claim 3, further comprising:

calculating a first error associated with the first prediction model;

calculating a second error associated with the second prediction model; and

calculating a third error associated with the third prediction model.

6. The method of claim 5, wherein the step of predicting the combined resource usage by applying a combination of the first and second prediction models is performed upon a first condition based on a comparison of the third error against a sum of the first and second errors.

7. The method of claim 6, wherein the step of predicting the combined resource usage by applying the third prediction model is performed upon a second different condition based on a comparison of the third error against the sum of the first and second errors.

8. The method of claim 1, wherein aggregating the first and second traces of resource utilization metrics of the first and second applications comprises:

aggregating each of the first traces of resource utilization metrics to each of the second traces of resource utilization metrics when the first and second traces are of the same resource utilization metric.

9. The method of claim 1, wherein:

generating the first prediction model comprises generating the first prediction model for each virtual machine guest domain and each driver domain in the virtual environment; and

generating the second prediction model comprises generating the second prediction model for each virtual machine guest domain and each driver domain in the virtual environment.

10. The method of claim 1, wherein:

generating the first prediction model comprises generating the first prediction model for each virtual machine guest domain and a hypervisor in the virtual environment; and

generating the second prediction model comprises generating the second prediction model for each virtual machine guest domain and the hypervisor in the virtual environment.

11. A system for optimizing a prediction of resource usage of multiple applications running in a virtual environment, comprising:

a first hardware platform implementing a native hardware system in which an application natively resides and executes, the native hardware system operating to execute a predetermined set of benchmarks that includes at least one of a network-intensive workload and a disk-intensive workload;

a second hardware platform implementing a virtual environment therein, the virtual environment operating to execute the predetermined set of benchmarks;

a third hardware platform operating to collect first resource usage traces from the first hardware platform that result from the execution of the predetermined set of benchmarks in the native hardware system and second resource usage traces from the second hardware platform that result from the execution of the predetermined set of benchmarks in the virtual environment;

wherein the third hardware platform further operating to generate at least one prediction model that predicts a resource usage of each of the multiple applications executing in the virtual environment based on the collected first and second resource usage traces; and

wherein the third hardware platform further operating to;

a) collect first application traces of resource utilization metrics in the native hardware system based on an execution of a first one of the multiple applications in the native hardware system;

b) collect second application traces of resource utilization metrics in the native hardware system based on an execution of a second one of the multiple applications in the native hardware system;

c) apply the at least one prediction model to the collected first application traces of the first application to calculate a first estimate resource usage in the virtual environment by the first application;

d) apply the at least one prediction model to the collected second application traces of the second application to calculate a second estimate resource usage in the virtual environment by the second application; and

e) predicting an overall resource usage in the virtual environment by the first and second applications by aggregating the first and second estimate resource usages.

12. The system of claim 11, wherein the third hardware platform further operating to:

generate a third prediction model that maps a first selected set of the first traces of a selected one of the first resource utilization metrics to the second traces of resource utilization metrics;

generate a fourth prediction model that maps a second different selected set of the first traces of the selected first resource utilization metric to the second traces of resource utilization metrics;

aggregate the first application traces of the first application and the second application traces of the second application into combined application traces of resource utilization metrics; and

predict an overall resource usage of the first and second applications running in the virtual environment by applying the first and second prediction models to different sets of the combined application traces of resource utilization metrics.

13. The system of claim 11, wherein the virtual environment includes a driver domain and a virtual machine (VM) guest domain, and the second resource usage traces comprises:

a percentage of processor utilization in the virtual environment that is attributed to a utilization of hardware processing resources in the second hardware platform by the driver domain; and

a percentage of processor utilization in the virtual environment that is attributed to a utilization of hardware processing resources in the second hardware platform by the VM guest domain.

14. The system of claim 11, wherein the virtual environment includes a hypervisor and a virtual machine (VM) guest domain, and the second resource usage traces comprises:

a percentage of processor utilization in the virtual environment that is attributed to a utilization of hardware processing resources in the second hardware platform by the hypervisor; and

a percentage of processor utilization in the virtual environment that is attributed to a utilization of hardware processing resources in the second hardware platform by the VM guest domain.

15. A computer readable medium on which is encoded programming code executed by a computerized system for:

providing a predetermined set of benchmarks;

executing the predetermined set of benchmarks in a native hardware system in which the application natively resides;

executing the predetermined set of benchmarks in the virtual environment;

collecting first traces of first resource utilization metrics in the native hardware system based on the execution of the predetermined set of benchmarks in the native hardware system;

collecting second traces of second resource utilization metrics in the virtual environment based on the execution of the predetermined set of benchmarks in the virtual environment;

generating a first prediction model that maps a first selected set of the first traces of a selected one of the first resource utilization metrics to the second traces of resource utilization metrics;

generating a second prediction model that maps a second selected set of the first traces of the selected first resource utilization metric to the second traces of resource utilization metrics;

collecting first application traces of resource utilization metrics in the native hardware system based on an execution of a first application in the native hardware system;

collecting second application traces of resource utilization metrics in the native hardware system based on an execution of a second application in the native hardware system;

aggregating the first application traces of the first application and the second application traces of the second application into combined application traces of resource utilization metrics; and

predicting a combined resource usage of the first and second applications running in the virtual environment by applying the first and second prediction models to the combined application traces of resource utilization metrics.