Proactive link load balancing to maintain quality of link
Disclosed embodiments describe systems and methods for predicting health of a link. A device in communication with a link can identify profile information of a stream of network traffic traversing the link. The device can determine a first prediction of health of the link by applying one or more rules to the plurality of parameters of the profile information. The device can determine a second prediction of health of the link by applying a classifier to one or more timed sequences of the plurality of parameters of the profile information. The device can establishes a respective weight for each of the first prediction of health and the second prediction of heath. The device can select, using the respective weight, between the first prediction of health and the second prediction of health to provide a predictor of the health of the link.
1. A method, comprising:
(a) identifying, by a device, a plurality of parameters about a link in communication with the device;
(b) applying, by the device, one or more rules to the plurality of parameters of the link to establish a first prediction of health of the link;
(c) determining, by a classifier trained using machine learning, a second prediction of health of the link based at least on the plurality of parameters, wherein the second prediction of health is a different type of prediction than the first prediction of health;
(d) selecting, by the device, one of the first prediction of health or the second prediction of health as a health of the link based at least on a respective weight of each of the first prediction of health and the second prediction of health; and
(e) reducing, by the device, a use of the link based at least on the health of the link.
2. The method of claim 1 , wherein (a) further comprises identifying, by the device, the plurality of parameters from network traffic traversing the device and communicated via the link.
3. The method of claim 1 , wherein the plurality of parameters comprises one or more of the following: latency, utilization, loss of packets, jitter, queue size, number of connections, and type of connection.
4. The method of claim 1 , wherein (b) further comprises applying, by the device, the one or more rules as a function of the plurality of parameters.
5. The method of claim 1 , wherein (c) further comprises applying, by the device, the classifier to timed sequences of the plurality of parameters.
6. The method of claim 5 , further comprising creating the one or more timed sequences of the plurality of parameters responsive to the health of the link one of climbing above or falling below a predetermined threshold.
7. The method of claim 1 , wherein (d) further comprises applying a function to the respective weight of the first prediction of health and the second prediction of health to select between the first prediction of health and the second prediction of health.
8. The method of claim 1 , further comprising determining, by the device responsive to comparing the health of the link to a threshold, that the link has one of high latency, high jitter or high loss.
9. The method of claim 8 , further comprising causing, by the device responsive to the determination, one or more connections in a queue of the device to use a different link.
10. The method of claim 8 , further comprising using, by the device responsive to the determination, a predetermined portion of bandwidth via the link.
11. A system comprising:
a device comprising one or more processors, coupled to memory and configured to:
identify a plurality of parameters about a link in communication with the device;
apply one or more rules to the plurality of parameters of the link to establish a first prediction of health of the link;
determine, via a classifier trained using machine learning, a second prediction of health of the link based at least on the plurality of parameters, wherein the second prediction of health is a different type of prediction of health than the first prediction of health;
select one of the first prediction of health or the second prediction of health as a health of the link based at least on a respective weight of each of the first prediction of health and the second prediction of health; and
reduce a use of the link based at least on the health of the link.
12. The system of claim 11 , wherein the device is further configured to identify the plurality of parameters from network traffic traversing the device and communicated via the link.
13. The system of claim 11 , wherein the plurality of parameters comprises one or more of the following: latency, utilization, loss of packets, jitter, queue size, number of connections, and type of connection.
14. The system of claim 11 , wherein the device is further configured to apply the one or more rules as a function of the plurality of parameters.
15. The system of claim 11 , wherein the device is further configured to apply the classifier to timed sequences of the plurality of parameters.
16. The system of claim 15 , wherein the device is further configured to creating the one or more timed sequences of the plurality of parameters responsive to the health of the link one of climbing above or falling below a predetermined threshold.
17. The system of claim 11 , wherein the device is further configured to apply a function to the respective weight of the first prediction of health and the second prediction of health to select between the first prediction of health and the second prediction of health.
18. The system of claim 11 , wherein the device is further configured to determine, responsive to comparing the health of the link to a threshold, that the link has one of high latency, high jitter or high loss.
19. The system of claim 18 , wherein the device is further configured to cause, responsive to the determination, one or more connections in a queue of the device to use a different link.
20. The system of claim 18 , wherein the device is further configured to use, responsive to the determination, a predetermined portion of bandwidth via the link.