Failover mechanism based on historical usage data

• US 8,831,202 B1
• Filed: 06/18/2012
• Issued: 09/09/2014
• Est. Priority Date: 01/23/2009
• Status: Active Grant

• Current Assignee: T-Mobile Innovations LLC (Deutsche Telekom AG)
• Original Assignee: Sprint Communications Company LP (Deutsche Telekom AG)
• Inventors: Abidogun, Ayodeji, Subramanyan, Badri P., Routt, William James, Davis, John W., Upadhyay, Piyush
• Primary Examiner(s): Addy, Thjuan K
• Application Number:
  US13/526,262
• Time in Patent Office: 813 Days
• Field of Search: 379/221.03, 379/221.04, 379/221.05, 379/219, 379/221.01
• US Class Current: 379/221.03
• CPC Class Codes: H04L 45/28 using route fault recovery

Abstract

Computer readable media and methods are provided for initiating a failover state in a network element. A failover is when the element'"'"'s traffic is routed to a redundant element which takes over the functions of the element, and any session state information stored in the element is also transferred to the redundant element. If a failover occurs with an upstream element, a downstream element is not aware of the failover and does not transfer its session state information to its redundant element. By maintaining a database of historic usage information, the downstream element can determine that an upstream failover has occurred when the amount of data it is processing drops below a minimum usage threshold. The minimum usage threshold is based on the historical usage data and a relationship between a given time in the past and the current time.

Claims

1. One or more nontransitory computer-readable media having computer-useable instructions embodied thereon for performing a method of initiating a failover state of an element in a telecommunications network, the method comprising:

determining an anticipated amount of usage for the network element at a present time;

based on said anticipated amount of usage, determining a minimum-usage threshold to be associated with said element for said present time, wherein said minimum-usage threshold defines a minimum value of a data-processing metric, below which said network element would be deemed to be in a fail status;

calculating said data-processing metric based on current processing being handled by said network element;

determining that said data-processing metric is below said minimum-usage threshold; and

initiating said failover state of said network element based on said determining that said data-processing metric is below said minimum-usage threshold.

2. The media of claim 1, wherein said anticipated amount of usage comprises one of a number of users and a volume of data.

3. The media of claim 1, wherein said initiating includes directing control over data that was previously being handled by said network element to a second network element.

4. The media of claim 3, wherein said data that was previously being handled by said network element comprises session state data.

5. One or more nontransitory computer-readable media having computer-useable instructions embodied thereon for determining when an upstream failure has occurred in a network, the method comprising:

determining an anticipated amount of usage associated with a network element at a current time;

determining a minimum usage threshold to be associated with said network element at said current time based on at least said anticipated amount of usage, below which minimum usage threshold said upstream failure would be deemed to have occurred;

providing a low-usage determination that an actual amount of usage associated with said network element at said current time is below said minimum usage threshold, and

determining that said upstream failure has occurred based on said low-usage determination.

6. The media of claim 5, wherein said amount of usage comprises one of a number of users and a volume of data.

7. The media of claim 5, wherein determining said anticipated amount of usage is further based on said one or more relevancy factors that indicate a measure of a relevancy of said data-processing metric to said current time.

8. The media of claim 7, wherein said one or more relevancy factors comprise one of time of day, day of week, day of month, a holiday, and size of a user base.

9. A method for initiating a failover state of an element in a telecommunications network, the method comprising:

determining a minimum-usage threshold associated with said element for a current time, wherein said minimum-usage threshold defines a minimum value of a current data-processing metric, below which said element would be deemed to be in a fail status;

providing a determination that an actual amount of usage for said element at said current time is below said minimum-usage threshold; and

initiating said failover state of said element based on said determination that said actual amount of usage for said element at said current time is below said minimum usage threshold.

10. The method of claim 9, wherein said current data-processing metric is based on at least one of a number of users and a volume of data.

11. The method of claim 9, wherein said initiating includes directing control over data that was previously being handled by said element to a second element.

12. The method of claim 9, further comprising determining a relevance of said data-processing measures to said current time.

13. The method of claim 12, wherein determining a minimum-usage threshold is further based on the relevance of said data-processing measures to said current time.

14. The method of claim 13, wherein said relevance of said data-processing measures to said current time is based on at least one of time of day, day of week, day of month, a holiday, and size of a user base.