System and method for automated criteria based deployment of virtual machines across a grid of hosting resources

• US 8,458,717 B1
• Filed: 04/09/2009
• Issued: 06/04/2013
• Est. Priority Date: 09/23/2008
• Status: Active Grant

• Current Assignee: Open Invention Network LLC
• Original Assignee: GoGrid LLC (Apollo Global Management, Inc.)
• Inventors: Carr, Jeffery, Keagy, John Martin, Lappas, Paul
• Primary Examiner(s): LI, AIMEE J
• Application Number:
  US12/421,599
• Time in Patent Office: 1,517 Days
• Field of Search: 718/104, 718/105
• US Class Current: 718/104
• CPC Class Codes: G06F 2009/45562 Creating, deleting, cloning virtual machine instances
• CPC Class Codes: G06F 8/63 Image based installation; Cloning; Build to order
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• CPC Class Codes: G06F 9/45537 Provision of facilities of other operating environments, e.g. WINE I/O emulation G06F13/105
• CPC Class Codes: G06F 9/4555 Para-virtualisation, i.e. guest operating system has to be modified
• CPC Class Codes: G06F 9/45558 Hypervisor-specific management and integration aspects

Abstract

Some embodiments provide a method. The method interfaces with several nodes of a hosting system to collect statistics that includes available resources of each node for hosting virtual machines. The method identifies a set of nodes with sufficient available resources to host a configuration for a virtual machine based on the collected statistics. The method determines an optimal node from the set of nodes that hosts the configuration with a particular set of performance parameters that yields improved performance than when hosting the configuration using any of the other nodes in the set of nodes. The method deploys the configuration for the virtual machine to the optimal node.

Claims

1. A method comprising:

interfacing with a plurality of nodes of a hosting system to collect statistics comprising available resources of each node for hosting virtual machines, wherein each node comprises one of a plurality of different hypervisors that provide different levels of performance for different virtual machine configurations;

identifying a set of nodes with sufficient available resources to host a particular user-specified configuration for a virtual machine based on the collected statistics;

determining an optimal node from the set of nodes by analyzing, for each particular node in the set of nodes, behavioral characteristics of the particular configuration and performance characteristics of the hypervisor operating on the particular node for the particular configuration, such that the optimal node that hosts the particular configuration yields improved performance than when hosting the particular configuration using any of the other nodes in the set of nodes; and

deploying the particular configuration for the virtual machine to the optimal node.

2. The method of claim 1, wherein said resources comprise at least one of processing resources, memory, disk space, and network input/output bandwidth of the plurality of nodes and said statistics comprise usage statistics for at least one of processing resources, memory, disk space, and network input/output bandwidth of the plurality of nodes.

3. The method of claim 1, wherein the hypervisor of the optimal node provides a particular level of performance that is greater than levels of performance provided by other hypervisors in the set of nodes operating with said configuration.

4. The method of claim 1, wherein the plurality of different hypervisors comprises a type 1 hypervisor that accesses hardware resources of a node through a host operating system and a type 2 hypervisor that directly accesses hardware resources of a node without a host operating system intermediary.

5. The method of claim 1, wherein the plurality of different hypervisors comprises hypervisors from a plurality of vendors.

6. The method of claim 1, wherein determining the optimal node further comprises selecting a node with an amount of available resources that most closely match resources required to host the particular configuration.

7. The method of claim 1, wherein said statistics identify the hypervisor operating at each node of the plurality of nodes.

8. A method comprising:

providing a plurality of nodes with resources for hosting a plurality of virtual machines, wherein each of the plurality of resource nodes comprises one of a plurality of hypervisors for provisioning sets of resources of the node amongst the plurality of virtual machines; and

providing a hypervisor management module for interfacing with the hypervisors on the plurality of resource nodes to optimally deploy the plurality of virtual machines across the plurality of resource nodes.

9. The method of claim 8, wherein the hypervisor management module for interfacing with the hypervisors is for retrieving statistics of resources from the nodes for facilitating deployment decisions of the virtual machines.

10. The method of claim 8, wherein the different hypervisors specify different messaging interfaces by which to communicate with each of the different hypervisors.

11. The method of claim 8, wherein optimally deploying the plurality of virtual machines comprises determining a node from a set of nodes which executes the configuration to provide a particular level of performance, wherein the particular level of performance is greater than the level of performance provided when executing the configuration on other nodes in the set of nodes.

12. The method of claim 8, wherein optimally deploying the plurality of virtual machines comprises determining a node from a set of nodes with a particular amount of available resources that most closely match an amount of resources required to host the configuration, wherein other nodes in the set of nodes comprise available resources that exceed the particular amount of available resources.

13. A computer readable hardware medium storing a hypervisor management module that interfaces with a plurality of nodes of a hosting system, the nodes comprising resources for hosting virtual machines of a plurality of different users, the hypervisor management module for execution by at least one processor of the hosting system, the hypervisor management module comprising:

a set of instructions for collecting, for each particular node in the hosting system, a set of statistics for the particular node, wherein a first set of statistics is collected from a utility management module operating as a virtual machine on the particular node and a second set of statistics is collected from a hypervisor on the particular node;

a set of instructions for normalizing the collected statistics from the hypervisors and utility management modules to produce uniformly formatted statistics;

a set of instructions for identifying a set of nodes with sufficient available resources to host a particular user-specified configuration for a virtual machine based on the normalized collected statistics;

a set of instructions for determining an optimal node from the set of nodes that hosts the particular configuration with a particular set of performance parameters that yields improved performance than when hosting the particular configuration using any of the other nodes in the set of nodes; and

a set of instructions for deploying the particular configuration for the virtual machine to the optimal node.

14. The computer readable hardware medium of claim 13, wherein said resources comprise at least one of processing resources, memory, disk space, and network input/output bandwidth of the plurality of nodes and said statistics comprise usage statistics for at least one of processing resources, memory, disk space, and network input/output bandwidth of the plurality of nodes.

15. The computer readable hardware medium of claim 13, wherein each node comprises one of a plurality of different hypervisors, each hypervisor for facilitating access between the configuration of the virtual machine and resources of a node with different levels of performance, wherein the set of instructions for determining the optimal node comprises a set of instructions for selecting a node from the set of nodes with a hypervisor that provides a particular level of performance when operating with said particular configuration, wherein the particular level of performance is greater than levels of performance provided by other hypervisors in the set of nodes operating with said particular configuration.

16. The computer readable hardware medium of claim 15, wherein the plurality of different hypervisors comprises a type 1 hypervisor that accesses hardware resources of a node through a host operating system and a type 2 hypervisor that directly accesses hardware resources of a node without a host operating system intermediary.

17. The computer readable hardware medium of claim 15, wherein the plurality of different hypervisors comprises hypervisors from a plurality of vendors.

18. The computer readable hardware medium of claim 13, wherein the set of instructions for determining the optimal node comprises a set of instructions for selecting a node with an amount of available resources that most closely match resources required to host the particular configuration.

19. The computer readable hardware medium of claim 13, wherein said statistics identify the hypervisor operating at each node of the plurality of nodes.