Method and system for managing tasks by dynamically scaling centralized virtual center in virtual infrastructure

• US 8,631,403 B2
• Filed: 01/04/2010
• Issued: 01/14/2014
• Est. Priority Date: 01/04/2010
• Status: Active Grant

• Current Assignee: Vmware LLC (Broadcom, Inc.)
• Original Assignee: VMware, Inc. (Broadcom, Inc.)
• Inventors: Soundararajan, Vijayaraghavan, Meng, Shicong
• Primary Examiner(s): Lee, Thomas
• Assistant Examiner(s): Chen, Xuxing
• Application Number:
  US12/651,598
• Publication Number:
  US 20110167421A1
• Time in Patent Office: 1,471 Days
• Field of Search: 718/1
• US Class Current: 718/1
• CPC Class Codes: G06F 2009/45562 Creating, deleting, cloning virtual machine instances
• CPC Class Codes: G06F 2009/4557 Distribution of virtual machine instances; Migration and load balancing
• CPC Class Codes: G06F 2009/45595 Network integration; Enabling network access in virtual machine instances
• CPC Class Codes: G06F 30/394 Routing G06F30/396 takes precedence
• CPC Class Codes: G06F 9/45533 Hypervisors; Virtual machine monitors
• 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
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system
• CPC Class Codes: G06F 9/5088 involving task migration
• CPC Class Codes: H04L 45/44 Distributed routing
• CPC Class Codes: H04L 47/125 by balancing the load, e.g. traffic engineering

Abstract

Methods, systems, and computer programs for performing management tasks in a virtual infrastructure are presented. The method includes detecting an increase, beyond a predetermined threshold, in the number of tasks waiting to be processed by a first virtual center (VC), which executes as a virtual machine (VM) in the virtual infrastructure. Further, the method includes operations for spawning one or more additional VCs executing as VMs in the virtual infrastructure and for distributing the managed objects handled by the first VC among the additional VCs. The tasks waiting to be processed are reallocated among the first VC and the additional VCs, and the reallocated tasks are then processed by the corresponding VC.

Claims

1. A method for performing management tasks in a virtual infrastructure that includes a plurality of hosts as managed objects, the method comprising:

detecting an increase, beyond a predetermined threshold, in a number of tasks waiting to be processed by a first virtual center (VC), the first VC executing as a virtual machine (VM) in the virtual infrastructure;

spawning one or more additional VCs executing as VMs in the virtual infrastructure, each first and additional VC being a management VM for the managed objects of the virtual infrastructure;

distributing the managed objects handled by the first VC among the additional VCs, including distributing the hosts such that each additional VC manages only a subset of the hosts;

reallocating the tasks waiting to be processed among the first VC and the additional VCs; and

processing the reallocated tasks by the corresponding VC.

2. The method as recited in claim 1, wherein data regarding hosts and VMs managed by the first VC and the additional VCs is stored in a shared global database.

3. The method as recited in claim 1, further including:

detecting persistent light management load on the first VC and additional VCs;

detecting VCs not needed for ongoing management of the virtual infrastructure;

determining which VCs that are not needed will be terminated;

reclaiming managed objects from the VCs that will be terminated; and

terminating the determined VCs.

4. The method as recited in claim 1, wherein detecting an increase further includes detecting the increase when a parameter at the first VC exceeds a threshold value, the parameter being selected from a group consisting of Central Processing Unit (CPU) use, memory use, number of tasks queued, number of hosts managed, or number of nodes managed.

5. The method as recited in claim 1, wherein spawning further includes waking up a suspended VC.

6. The method as recited in claim 1, wherein spawning further includes creating a VC by cloning a template VC.

7. The method as recited in claim 1, further including:

determining a number of additional VCs to be spawned based on management workload, processing speed of each additional VC, and a cost of spawning a new VC.

8. The method as recited in claim 1, wherein reallocating the tasks further includes transferring a number of M_o tasks to each additional VC, wherein M_o is equal to

9. The method as recited in claim 1, wherein spawning one or more additional VCs includes spawning one additional VC, wherein one additional VC is spawned each time the increase is detected in the first VC or in already spawned VCs.

10. The method as recited in claim 1, wherein detecting the increase further includes:

forecasting an increase in the number of tasks waiting to be processed, wherein one or more additional VCs are spawned to anticipate the forecasted increase, wherein the hosts execute application VMs and are managed by the VCs which are management VMs.

11. The method as recited in claim 10, wherein the forecasting includes estimating a number of application VMs being spawned at a beginning of a work day.

12. A system for dynamically scaling management infrastructure in a virtual infrastructure, the system comprising:

a plurality of hosts in the virtual infrastructure, wherein the hosts execute application VMs;

a first virtual center (VC) executing as a virtual machine (VM) in one of the hosts, the first VC being a management VM configured to perform management tasks for operation of a plurality of managed hosts of the plurality of hosts;

a master management node defined to detect an increase, beyond a predetermined threshold, in a number of management tasks waiting to be processed by the first VC for managing operations of the managed hosts, wherein upon detecting the increase the master management node performs operations of,

spawning one or more additional VCs executing as VMs in the virtual infrastructure, each of the additional VCs being management VMs configured to perform the management tasks, wherein each of the first and additional VCs is assigned to correspond to only a subset of the managed hosts, and

reallocating the management tasks waiting to be processed among the first VC and the additional VCs,

wherein each reallocated management task is processed by the VC that is assigned to correspond to the managed host to which the reallocated management task is related; and

a shared global database holding data regarding hosts and VMs managed by the first VC and the additional VCs.

13. The system as recited in claim 12, further including:

a client, wherein the client has a request for a managed object, wherein the master node provides information regarding which VC can satisfy the request for the client.

14. The system as recited in claim 12, wherein the shared global database includes an instruction translator to translate requests from a first database format to a second database format.

15. The system as recited in claim 12, wherein the first VC and the spawned additional VC are part of a same cluster.

16. The system as recited in claim 12, wherein reallocating the management tasks further includes transferring a number M_o of tasks to each additional VC,

17. A computer program embedded in a non-transitory computer-readable storage medium, when executed by one or more processors, for performing management tasks in a virtual infrastructure that includes a plurality of managed objects, the computer program comprising:

program instructions for detecting an increase, beyond a predetermined threshold, in a number of hosts or VMs being managed by a first virtual center (VC), the first VC executing as a virtual machine (VM) in the virtual infrastructure;

program instructions for spawning one or more additional VCs executing as VMs in the virtual infrastructure, each first and additional VC being a management VM for the managed objects of the virtual infrastructure;

program instructions for distributing managed objects handled by the first VC among the additional VCs, including distributing the hosts such that each additional VC manages only a subset of the hosts;

program instructions for reallocating the tasks waiting to be processed among the first VC and the additional VCs; and

program instructions for processing the reallocated tasks by the corresponding VC.

18. The computer program embedded in the non-transitory computer-readable storage medium as recited in claim 17, further including:

program instructions for segmenting tasks processed by the spawned VCs according to a task category.

19. The computer program embedded in the non-transitory computer-readable storage medium as recited in claim 17, wherein the program instructions for spawning one or more additional VCs further includes:

program instructions for creating a clone of the first VC.

20. The computer program embedded in the non-transitory computer-readable storage medium as recited in claim 17, further including:

program instructions for detecting persistent light management load on the first VC and additional VCs and for detecting VCs not needed for ongoing management of the virtual infrastructure;

program instructions for determining which VCs will be terminated;

program instructions for reclaiming managed objects from the VCs that will be terminated; and

program instructions for terminating the determined VCs.