Distributed and learning machine-based approach to gathering localized network dynamics

• US 10,425,294 B2
• Filed: 01/27/2014
• Issued: 09/24/2019
• Est. Priority Date: 01/06/2014
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Vasseur, Jean-Philippe, Mermoud, Gregory, Dasgupta, Sukrit
• Primary Examiner(s): Rifkin, Ben M
• Application Number:
  US14/164,444
• Publication Number:
  US 20150195144A1
• Time in Patent Office: 2,066 Days
• Field of Search: None
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
• CPC Class Codes: G06N 5/025 Extracting rules from data
• CPC Class Codes: G06N 99/00 Subject matter not provided for in other groups of this subclass
• CPC Class Codes: H04L 41/12 Discovery or management of network topologies
• CPC Class Codes: H04L 41/5025 by proactively reacting to service quality change, e.g. by reconfiguration after service quality degradation or upgrade
• CPC Class Codes: H04L 43/0852 Delays
• CPC Class Codes: H04L 45/02 Topology update or discovery
• CPC Class Codes: H04L 45/70 Routing based on monitoring results
• CPC Class Codes: H04L 47/122 by diverting traffic away from congested entities
• CPC Class Codes: H04L 47/365 Dynamic adaptation of the packet size
• CPC Class Codes: H04W 24/02 Arrangements for optimising operational condition
• CPC Class Codes: H04W 24/04 Arrangements for maintaining operational condition
• CPC Class Codes: Y04S 40/00 Systems for electrical power generation, transmission, distribution or end-user application management characterised by the use of communication or information technologies, or communication or information technology specific aspects supporting them

Abstract

In one embodiment, one or more reporting nodes are selected to report network metrics in a network. From a monitoring node in the network, a trigger message is sent to the one or more reporting nodes. The trigger message may trigger the one or more reporting nodes to report one or more network metrics local to the respective reporting node. In response to the trigger message, a report of the one or more network metrics is received at the monitoring node from one of the one or more reporting nodes.

Claims

1. A method, comprising:

generating, by a learning machine (LM), a list of one or more network metrics needed at the LM, wherein the list of one or more metrics are temporary localized network dynamics that are required by the LM to build a predictive model;

selecting, by the LM, one or more reporting nodes to report the one or more network metrics required by the LM;

generating, by the LM, a trigger message to be sent to the selected one or more reporting nodes, wherein the trigger message includes the list of one or more network metrics and a rule indicating that when the local condition is determined to have occurred locally at the selected one or more reporting nodes, the one or more network metrics are to be reported by the selected one or more reporting nodes to the LM and wherein the trigger message instructs the selected one or more reporting nodes to start reporting the list of one or more network metrics when the local condition occurs;

sending, from the LM, the trigger message to the selected one or more reporting nodes to establish the rule on the selected one or more reporting nodes locally; and

once the local condition occurs on the selected one or more reporting nodes, receiving, at the LM, a report including the one or more network metrics from one of the selected one or more reporting nodes, wherein ephemeral network behaviors occurring locally on the selected one or more reporting nodes are captured in the report.

2. The method according to claim 1, further comprising:

including an indication of the one or more network metrics in the trigger message, wherein

the one or more reported network metrics correspond to the one or more network metrics.

3. The method according to claim 1, wherein the one or more network metrics are selected for reporting by the selected one or more reporting nodes based on the trigger message.

4. The method according to claim 1, further comprising:

defining a rule indicating when the one or more network metrics are to be reported by the selected one or more reporting nodes; and

including an indication of the defined rule in the trigger message, wherein

the one or more network metrics are reported by the selected one or more reporting nodes in accordance with the defined rule.

5. The method according to claim 4, wherein:

the rule specifies a condition of whether a network metric local to a reporting node changes during a period of time, and

when the network metric changes during the period of time, the network metric is reported by the reporting node.

6. The method according to claim 4, wherein the rule specifies a threshold linked to the one or more network metrics.

7. The method according to claim 4, wherein:

the rule specifies a first threshold associated with a first condition and a second threshold associated with a second condition.

8. The method according to claim 1, further comprising:

defining a frequency at which the selected one or more reporting nodes are to report the one or more network metrics; and

including an indication of the defined frequency in the trigger message, wherein

reports of the one or more network metrics are received at a frequency corresponding to the defined frequency.

9. The method according to claim 1, further comprising:

defining a duration for which the trigger message is to remain active; and

including an indication of the defined duration in the trigger message, wherein

the report of the one or more network metrics is received only while the trigger message is active.

10. The method according to claim 1, wherein the one or more network metrics improve training efficiency of the LM.

11. An apparatus, comprising:

one or more network interfaces that communicate with a network;

a processor coupled to the one or more network interfaces and configured to execute a process; and

a memory configured to store program instructions which contain the process executable by the processor, the process comprising;

generating, as a learning machine, (LM), a list of one or more network metrics needed at the LM, wherein the list of one or more metrics are temporary localized network dynamics that are required by the LM to build a predictive model;

selecting, as the LM, one or more reporting nodes to report one or more network metrics in the network required by the LM;

generating, as the LM, a trigger message to be sent to the selected one or more reporting nodes, wherein the trigger message includes the list of one or more network metrics and a rule indicating that when the local condition is determined to have occurred locally at the selected one or more reporting nodes, the one or more network metrics are to be reported by the selected one or more reporting nodes to the LM and wherein the trigger message instructs the selected one or more reporting nodes to start reporting the list of one or more network metrics when the local condition occurs;

sending, as the LM in the network, the trigger message to the selected one or more reporting nodes to establish the rule on the one or more reporting nodes locally; and

once the local condition occurs on the selected one or more reporting nodes, receiving, at the LM, a report includes the one or more network metrics from one of the selected one or more reporting nodes, wherein ephemeral network behaviors occurring locally on the selected one or more reporting nodes are captured in the report.

12. The apparatus according to claim 11, wherein the process further comprises:

including an indication of the one or more network metrics in the trigger message, wherein

the one or more reported network metrics correspond to the one or more network metrics.

13. The apparatus according to claim 11, wherein the one or more network metrics are selected for reporting by the selected one or more reporting nodes based on the trigger message.

14. The apparatus according to claim 11, wherein the process further comprises:

defining a rule indicating when the one or more network metrics are to be reported by the selected one or more reporting nodes; and

including an indication of the defined rule in the trigger message, wherein

the one or more network metrics are reported by the selected one or more reporting nodes in accordance with the defined rule.

15. The apparatus according to claim 14, wherein:

the rule specifies a condition of whether a network metric local to a reporting node changes during a period of time, and

when the network metric changes during the period of time, the network metric is reported by the reporting node.

16. The apparatus according to claim 14, wherein the rule specifies a threshold.

17. The apparatus according to claim 14, wherein:

the rule specifies a first threshold associated with a first condition and a second threshold associated with a second condition.

18. The apparatus according to claim 11, wherein the process further comprises:

defining a frequency at which the selected one or more reporting nodes are to report the one or more network metrics; and

including an indication of the defined frequency in the trigger message, wherein

reports of the one or more network metrics are received at a frequency corresponding to the defined frequency.

19. The apparatus according to claim 11, wherein the process further comprises:

defining a duration for which the trigger message is to remain active; and

including an indication of the defined duration in the trigger message, wherein

the report of the one or more network metrics is received only while the trigger message is active.

20. The apparatus according to claim 11, wherein the one or more network metrics improve training efficiency of the LM.

21. A tangible non-transitory computer readable medium storing program instructions that cause a computer to execute a process, the process comprising:

generating, as a learning machine, (LM), a list of one or more network metrics needed at the LM, wherein the list of one or more metrics are temporary localized network dynamics that are required by the LM to build a predictive model;

selecting, as the LM, one or more reporting nodes to report one or more network metrics required by the LM;

generating, as the LM, a trigger message to be sent to the selected one or more reporting nodes, wherein the trigger message includes the list of one or more network metrics and a rule indicating that when the local condition is determined to have occurred locally at the selected one or more reporting nodes, the one or more network metrics are to be reported by the selected one or more reporting nodes to the LM and wherein the trigger message instructs the selected one or more reporting nodes to start reporting the list of one or more network metrics when the local condition occurs;

sending, as the LM, the trigger message to the one or more reporting nodes to establish the rule on the selected one or more reporting nodes locally; and

once the local condition occurs on the selected one or more reporting nodes, receiving, as the LM, a report includes the one or more network metrics from one of the selected one or more reporting nodes, wherein ephemeral network behaviors occurring locally on the selected one or more reporting nodes are captured in the report.