Starting up at least one virtual machine in a physical machine by a load balancer

• US 8,185,893 B2
• Filed: 10/27/2006
• Issued: 05/22/2012
• Est. Priority Date: 10/27/2006
• Status: Active Grant

• Current Assignee: Regional Resources Ltd (Imperial Metals Corp.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Hyser, Chris D., McKee, Bret A.
• Primary Examiner(s): To, Jennifer
• Assistant Examiner(s): KUMABE, BLAKE K
• Application Number:
  US11/588,635
• Publication Number:
  US 20080104608A1
• Time in Patent Office: 2,034 Days
• Field of Search: 718/1, 718/106
• US Class Current: 718/1
• CPC Class Codes: G06F 2009/4557 Distribution of virtual machine instances; Migration and load balancing
• CPC Class Codes: G06F 2209/5015 Service provider selection
• CPC Class Codes: G06F 2209/5022 Workload threshold
• CPC Class Codes: G06F 2209/508 Monitor
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals
• CPC Class Codes: G06F 9/5088 involving task migration
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

A system has plural physical machines that contain virtual machines. A load balancer receives a request from a client. In response to the request, it is determined whether at least one additional virtual machine should be started up. In response to determining that at least one additional virtual machine should be started up, the load balancer sends at least one command to start up the at least one additional virtual machine in at least one of the physical machines.

Claims

1. A method for use in a system having plural physical machines that each contains one or more active virtual machines, comprising:

receiving, at a load balancer, a request from a client;

in response to the request, determining whether at least one additional virtual machine should be started up;

in response to determining that at least one additional virtual machine should be started up, the load balancer sending at least one command to start up the at least one additional virtual machine in at least one of the physical machines;

a placement controller modifying placement of the active virtual machines on the physical machines in response to starting up of the at least one additional virtual machine;

computing by the placement controller indicators associated with corresponding plural different layouts of the active virtual machines and the at least one additional virtual machine on the physical machines;

selecting, by the placement controller, one of the plural layouts based on the indicators;

placement controller one of the plural layouts based on the indicators;

determining, by the load balancer, whether a workload loading of the active virtual machines and the at least one additional virtual machine has fallen below a threshold; and

in response to determining that the workload loading has fallen below the threshold, disabling a particular one of the active virtual machines and the at least one additional virtual machine.

2. The method of claim 1, wherein sending the at least one command to start up the at least one additional virtual machine comprises sending the at least one command to activate the at least one additional virtual machine that is in a suspend state.

3. The method of claim 1, wherein sending the at least one command to start up the at least one additional virtual machine comprises sending the at least one command to create the at least one additional virtual machine.

4. The method of claim 1, further comprising:

the load balancer balancing workload of the active virtual machines and the at least one additional virtual machine in the physical machines, wherein the workload is responsive to requests received over a network from one or more clients.

5. The method of claim 1, further comprising:

the placement controller modifying placement of the active virtual machines and the at least one additional virtual machine on the physical machines in response to a changing condition.

6. The method of claim 1, wherein disabling the particular virtual machine comprises one of turning off the particular virtual machine and placing the particular virtual machine into a low power mode.

7. The method of claim 6, wherein disabling the particular virtual machine is performed in response to a command from the load balancer.

8. The method of claim 6, wherein disabling the particular virtual machine is performed under control of the placement controller that manages placement of virtual machines on physical machines.

9. The method of claim 6, wherein determining that the at least one additional virtual machine should be started up is in response to load conditions of the active virtual machines on the plural physical machines.

10. The method of claim 9, further comprising:

receiving, by the load balancer, a second request;

determining whether the active virtual machines are able to handle the second request; and

in response to determining that the active virtual machines are able to handle the second request, sending the second request to an identified one of the active virtual machines.

11. The method of claim 1, further comprising:

redistributing the workload of the active virtual machines and the at least one additional virtual machine such that at least one of the active virtual machines and the at least one additional virtual machine is idle; and

selecting the idle virtual machine as the particular virtual machine to disable.

12. An article comprising a computer-readable storage medium storing instructions that when executed cause a system having a processor to:

receive a request for a service to be performed by an arrangement of physical machines containing active virtual machines;

determine whether the active virtual machines in the physical machines are able to handle the request;

in response to determining that the active virtual machines are unable to handle the request, start up at least one additional virtual machine to handle the request;

determine whether a workload loading of the active virtual machines and the at least one additional virtual machine has fallen below a threshold;

in response to determining that the workload loading has fallen below the threshold, disable a particular one of the active virtual machines and the at least one additional virtual machine;

migrate virtual machines between or among physical machines to achieve a predefined policy;

compute indicators associated with corresponding plural different layouts of the active virtual machines and the at least one additional virtual machine on the physical machines, wherein the indicators provide information regarding performances of corresponding layouts, and wherein each of the indicators is computed based on parameters associated with a corresponding one of the plural layouts;

compare the indicators; and

select one of the plural layouts based on the comparing.

13. The article of claim 12, wherein starting up the at least one virtual machine comprises one of (1) activating the at least one additional virtual machine from a suspend state, and (2) creating the at least one additional virtual machine.

14. The article of claim 12, wherein determining that the active virtual machines are unable to handle the request comprises detecting that an average workload per active virtual machine has exceeded a predefined threshold.

15. The article of claim 12, wherein the instructions when executed cause the system to further:

redistribute the workload of the active virtual machines and the at least one additional virtual machine such that at least one of the active virtual machines and the at least one additional virtual machine is idle; and

select the idle virtual machine as the particular virtual machine to disable.

16. A method for use in a system having plural physical machines that contain active virtual machines, comprising:

receiving, at a load balancer, a request from a client;

in response to the request, determining whether at least one additional virtual machine should be started up;

in response to determining that at least one additional virtual machine should be started up, the load balancer sending at least one command to start up the at least one additional virtual machine in at least one of the physical machines;

determining, by the load balancer, whether a workload loading of the active virtual machines and the at least one additional virtual machine has fallen below a threshold;

in response to determining that the workload loading has fallen below the threshold, disabling a particular one of the active virtual machines and the at least one additional virtual machine;

a placement controller selecting placement of the active virtual machines and the at least one additional virtual machines on the physical machines to achieve a predefined policy;

computing, by the placement controller, indicators associated with corresponding plural different layouts of the active virtual machines and the at least one additional virtual machine on the physical machines, where the indicators provide information regarding performances of the corresponding layouts, and wherein each of the indicators is computed based on parameters associated with a corresponding one of the plural layouts;

comparing, by the placement controller, the indicators; and

selecting, by the placement controller, one of the plural layouts based on the comparing.

17. A system comprising:

physical machines that each contains one or more virtual machines;

a request server comprising;

a processor; and

a load balancer executable on the processor to;

receive requests from one or more clients;

distribute requests across active virtual machines in the physical machines;

start up additional virtual machines in response to a workload of the active virtual machines exceeding a first threshold;

determine whether a workload loading of the active virtual machines and the additional virtual machines has fallen below a second threshold; and

in response to determining that the workload loading has fallen below the second threshold, disable a particular one of the active virtual machines and the additional virtual machines; and

a placement controller to;

modify placement of the active virtual machines on the physical machines in response to starting up of the additional virtual machines;

compute indicators associated with corresponding plural different layouts of the active virtual machines and the additional virtual machines on the physical machines; and

select one of the plural layouts based on the indicators.

18. The system of claim 17, wherein the request server further comprises a network monitor executable on the processor to receive workload information from the active virtual machines for use by the load balancer.