FAST LEARNING TO TRAIN LEARNING MACHINES USING SMART-TRIGGERED REBOOT

• US 20140223155A1
• Filed: 06/25/2013
• Published: 08/07/2014
• Est. Priority Date: 02/05/2013
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Vasseur, Jean-Philippe, Mermoud, Grgory, Dasgupta, Sukrit, Hui, Jonathan W.
• Granted Patent: US 9,563,440 B2
• Time in Patent Office: Days
• US Class Current: 713/2
• CPC Class Codes: G06F 9/4405 Initialisation of multiprocessor systems

Abstract

In one embodiment, a triggered reboot of a field area router (FAR) of a computer network is initiated, and gathered states of the FAR are saved. The nodes in the computer network are informed of the triggered reboot, and then feedback may be collected from the nodes in response to the triggered reboot. As such, it can be determined whether to complete the triggered reboot based on the feedback, and the FAR is rebooted in response to determining to complete the triggered reboot. In another embodiment, a node receives information about the initiated triggered reboot of the FAR, and determines whether it has critical traffic. If not, the node buffers non-critical traffic and indicates positive feedback in response to the triggered reboot, but if so, then the node continues to process the critical traffic and indicates negative feedback in response to the triggered reboot.

Claims

1. A method, comprising:

initiating a triggered reboot of a field area router (FAR) of a computer network;

saving gathered states of the FAR;

informing nodes in the computer network of the triggered reboot;

collecting feedback from the nodes in response to the triggered reboot;

determining whether to complete the triggered reboot based on the feedback; and

rebooting the FAR in response to determining to complete the triggered reboot.

2. The method as in claim 1, further comprising:

receiving a time to perform the triggered reboot.

3. The method as in claim 2, wherein the time is based on a predicted quiet period of the computer network.

4. The method as in claim 1, further comprising:

saving one or more of the gathered states by sending the gathered states to a network management server (NMS); and

restoring the one or more of the gathered states from the NMS after rebooting the FAR.

5. The method as in claim 1, further comprising:

saving one or more of the gathered states locally on the FAR.

6. The method as in claim 1, wherein informing the nodes comprises a broadcast message.

7. The method as in claim 1, wherein informing the nodes includes one or both of a time of the triggered reboot and an approximate duration of the triggered reboot.

8. The method as in claim 1, wherein the feedback is selected from positive feedback, negative feedback, and both positive and negative feedback.

9. The method as in claim 1, wherein determining whether to complete the triggered reboot is based on one of either a number or ratio of nodes indicating negative feedback.

10. The method as in claim 1, wherein determining whether to complete the triggered reboot is based on one of either a number or ratio of critical nodes of the nodes in the network indicating negative feedback.

11. The method as in claim 1, further comprising:

not rebooting the FAR in response to determining not to complete the triggered reboot; and

in response, sending a notification that the FAR is not rebooting.

12. The method as in claim 11, wherein determining not to complete the triggered reboot comprises determining to delay the triggered reboot.

13. A method, comprising:

receiving, at a node, information about an initiated triggered reboot of a field area router (FAR) of a computer network;

determining whether the node has critical traffic; and

in response to the node having no critical traffic, buffering non-critical traffic and indicating positive feedback in response to the triggered reboot; and

in response to the node having critical traffic, continuing to process the critical traffic and indicating negative feedback in response to the triggered reboot.

14. The method as in claim 13, wherein critical traffic is one or both of current critical traffic being processed or future expected critical traffic.

15. An apparatus, comprising:

one or more network interfaces to communicate with a computer network;

a processor coupled to the network interfaces and adapted to execute one or more processes; and

a memory configured to store a process executable by the processor, the process when executed operable to;

initiate a triggered reboot of a field area router (FAR) of the computer network;

save gathered states of the FAR;

inform nodes in the computer network of the triggered reboot;

collect feedback from the nodes in response to the triggered reboot;

determine whether to complete the triggered reboot based on the feedback; and

reboot the FAR in response to determining to complete the triggered reboot.

16. The apparatus as in claim 15, wherein the process when executed is further operable to:

receive a time to perform the triggered reboot.

17. The apparatus as in claim 15, wherein the process when executed is further operable to:

save the gathered states either locally on the FAR or by sending the gathered states to a network management server (NMS) and restoring the one or more of the gathered states from the NMS after rebooting the FAR.

18. The apparatus as in claim 15, wherein determining whether to complete the triggered reboot is based on one of:

a number of nodes indicating negative feedback, a ratio of nodes indicating negative feedback, a number of critical nodes of the nodes in the network indicating negative feedback, or a ratio of critical nodes indicating negative feedback.

19. The apparatus as in claim 15, wherein the process when executed is further operable to:

not reboot the FAR in response to determining not to complete the triggered reboot; and

in response, send a notification that the FAR is not rebooting.

20. An apparatus, comprising:

one or more network interfaces to communicate with a computer network;

a processor coupled to the network interfaces and adapted to execute one or more processes; and

a memory configured to store a process executable by the processor, the process when executed operable to;

receive, as a node, information about an initiated triggered reboot of a field area router (FAR) of the computer network;

determine whether the node has critical traffic;

in response to the node having no critical traffic, buffer non-critical traffic and indicating positive feedback in response to the triggered reboot; and

in response to the node having critical traffic, continue to process the critical traffic and indicating negative feedback in response to the triggered reboot.