Disaster recovery and automatic relocation of cloud services

• US 8,639,793 B2
• Filed: 03/04/2011
• Issued: 01/28/2014
• Est. Priority Date: 10/29/2010
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Kapur, Sukhdev S., Spiegel, Ethan M., Banerjee, Subrata, Singh, Sumeet, Ganesan, Ashok
• Primary Examiner(s): NGUYEN, PHUOC H
• Application Number:
  US13/040,629
• Publication Number:
  US 20120110186A1
• Time in Patent Office: 1,061 Days
• Field of Search: 709/220, 709/223, 709/226, 709/229
• US Class Current: 709/223
• CPC Class Codes: G06F 11/2025 using centralised failover control functionality
• CPC Class Codes: G06F 11/203 using migration
• CPC Class Codes: G06F 9/5072 Grid computing

Abstract

Techniques are provided to move the services performed on one device to another device in a cloud computing system for a variety of reasons including failure, maintenance or upgrade of the device. A notification is received that services performed by an impacted device in a domain of a plurality of hierarchical domains need to be moved. A determination is made as to whether there are replacement resources available in the domain to perform the services, and if so, the replacement resources are automatically rendered to perform the services. The process continues to higher level domains that have a view into the capabilities of subordinate domains in order to determine where to move the services within the cloud computing system.

Claims

1. A method comprising:

in a cloud computing system comprising a plurality of hierarchical levels including a pod hierarchical level, a data center hierarchical level that comprises a plurality of pods and a network hierarchical level that comprises a plurality of data centers, wherein each pod comprises a plurality of devices, receiving a notification at a pod management device in a first pod in a first data center, wherein the pod management device performs pod hierarchical level resource management operations for the first pod, the notification indicating that services associated with one or more impacted devices in the first pod need to be moved;

determining whether there are one or more replacement devices available in the first pod to perform the services of the one or more impacted devices;

when it is determined that there are one or more replacement devices in the first pod available to perform the services of the one or more impacted devices, automatically provisioning one or more replacement devices in the first pod to perform the services of the one or more impacted devices;

when it is determined that there are no replacement devices in the first pod available to perform the services of the one or more impacted devices, sending a request to a data center management device at the data center hierarchical level;

receiving the request at the data center management device of the first data center;

wherein the data center management device automatically determines whether there are one or more replacement devices available in the plurality of pods other than in the first pod; and

when it is determined that there are one or more replacement devices in a second pod available to perform the services of the one or more impacted devices, automatically provisioning the one or more replacement devices in the second pod to perform the services of the one or more impacted devices.

2. The method of claim 1, wherein the notification indicates that the one or more impacted devices in the first pod have failed.

3. The method of claim 1, where the notification indicates that the one or more impacted devices in the first pod will become unavailable due to maintenance.

4. The method of claim 1, wherein provisioning comprises rendering at least a portion of policies of the one or more impacted devices to the one or more replacement devices.

5. The method of claim 4, when it is determined that there are no replacement devices available in any other pods of the first data center, sending a request to a management device that performs network hierarchical level center resource management operations for the plurality of data centers.

6. The method of claim 5, and further comprising:

receiving the request at the management device that performs network hierarchical level center resource management operations for the plurality of data centers; and

determining whether there are one or more replacement devices available in other data centers that can perform the services of the one or more impacted devices in the first pod.

7. The method of claim 6, and further comprising forwarding a request to a management device that performs data center hierarchical level resource management operations for one of the plurality of data centers determined to have one or more replacement devices available to perform the services of the one or more impacted devices of the first pod.

8. The method of claim 7, and further comprising:

at the management device that performs network hierarchical level resource management operations for the plurality of data centers, receiving a notification that one of the plurality of data centers needs to be moved; and

determining whether another one or more of the plurality of data centers has resources available to handle the operations of the data center that needs to be moved.

9. The method of claim 8, and further comprising:

determining an order of priority for a plurality of virtual data center operations of the data center that needs to be moved; and

generating requests to move the plurality of virtual data center operations to one or more other data centers according to the order of priority so that higher priority virtual data center operations are moved before lower priority virtual data center operations.

10. The method of claim 7, and further comprising automatically sending the requests to the management device that performs the data center hierarchical level resource management operations for the one or more data centers.

11. The method of claim 1, wherein receiving the notification at the management device in the first pod is during a provisioning process of devices in the first pod.

12. The method of claim 1, wherein automatically provisioning is performed without human interaction or involvement.

13. A method comprising:

in a cloud computing system comprising a plurality of data centers each comprising a plurality of devices, receiving a notification at a management device that performs resource management operations for the plurality of data centers, the notification indicating that operations of a data center need to be moved;

determining whether another one or more of the plurality of data centers has a corresponding management device that manages resources in the one or more data centers that are available to satisfy the operations of the data center that need to be moved;

determining an order of priority for a plurality of virtual data center operations of the data center that need to be moved; and

generating requests to move the plurality of virtual data center operations to one or more other data centers according to the order of priority so that higher priority virtual data center operations are moved before lower priority virtual data center operations.

14. The method of claim 13, and further comprising automatically generating and sending a request to another one or more of the plurality of data centers determined to have resources available to satisfy the operations of the failed data center, wherein the request is configured to cause the corresponding management device that performs resource management operations for the other one or more data centers to automatically render the devices to perform the operations of the failed data center.

15. The method of claim 13, and further comprising automatically sending the requests to the management device that performs resource management operations for the one or more data centers.

16. An apparatus comprising:

a network interface device configured to enable communications over a network;

a processor coupled to the network interface device, the processor configured to;

receive a notification that services performed by an impacted device in a domain of a plurality of hierarchical domains of a cloud computing system need to be moved;

determine whether there is one or more replacement devices available in the domain to perform the services of the impacted device;

when it is determined that there are one or more replacement devices available in the domain, automatically provision the one or more replacement device in the domain to perform the services of the impacted device;

when it is determined that there are no replacement resources available in the domain to perform the services, automatically send a request to a management device that performs resource management operations at a higher hierarchical domain to identify other domains that may have resources available to perform the services;

when it is determined that there are resources available to perform the services in other domains, automatically provision the resources to perform the services of the impacted device.

17. The apparatus of claim 16, wherein instructions that are operable to automatically provision comprise instructions operable to render at least a portion of policies of the one or more impacted devices to the one or more replacement devices.

18. The apparatus of claim 16, wherein the instructions that are operable to receive comprise instructions operable to receive the notification indicating that the impacted device has failed.

19. The apparatus of claim 16, wherein the instructions that are operable to receive comprise instructions operable to receive the notification indicating that the services performed by the impacted device need to be moved for maintenance or upgrade purposes.

20. One or more non-transitory computer readable storage media encoded with instructions that when executed are operable to:

receive a notification at a management device that performs resource management operations for a plurality of data centers in a cloud computing system, the notification indicating that operations of a data center need to be moved;

determine whether another one or more of the plurality of data centers has a corresponding management device that manages resources in the one or more data centers that are available to satisfy the operations of the data center that need to be moved;

determine an order of priority for a plurality of virtual data center operations of the data center that need to be moved; and

generate requests to move the plurality of virtual data center operations to one or more other data centers according to the order of priority so that higher priority virtual data center operations are moved before lower priority virtual data center operations.

21. The non-transitory computer readable storage media of claim 20, and further comprising instructions that are operable to automatically generate and send a request to another one or more of the plurality of data centers determined to have resources available to satisfy the operations of the failed data center, wherein the request is configured to cause the corresponding management device that performs resource management operations for the other one or more data centers to automatically render the devices to perform the operations of the failed data center.