Methods and system for adaptively managing the distribution of network traffic
Embodiments of methods and systems for managing traffic across a WAN are disclosed. A method involves advertising an input rate limit across a WAN from a first node, advertising an input distribution array across the WAN from the first node, wherein the input distribution array includes forwarding class-specific weights, and adapting at least one of the input rate limit and the input distribution array in response to an error between a target forwarding class distribution and an observed input traffic distribution.
1. A method for managing traffic across a wide area network (WAN), the method comprising:
advertising an input rate limit across a WAN from a first node;
advertising an input distribution array across the WAN from the first node, wherein the input distribution array includes forwarding class-specific weights; and
adapting at least one of the input rate limit and the input distribution array in response to an error between a target forwarding class distribution and an observed input traffic distribution;
wherein adapting the input distribution array involves calculating an error as:
E=RMSE(Wn, Dn),
where,
i=integer set of forwarding classes with i being an integer greater than 1,
a normalized array of weights, Wn=w 0 /max(W); . . . w i /max(W), where max(W) is the highest weight, w, amongst the weights of the set of forwarding classes, and
a normalized traffic distribution, Dn=r 0 /max(R), . . . r i max(R), where max(R) is the highest rate, r, amongst the rates of the set of forwarding classes.
2. The method of claim 1 , wherein adapting the input distribution array involves incrementing a forwarding class-specific weight of the input distribution array.
3. The method of claim 1 , wherein adapting the input distribution array involves decrementing a forwarding class-specific weight of the input distribution array.
4. The method of claim 1 , further comprising identifying the forwarding class with the highest individual error and adapting the corresponding forwarding class-weight.
5. The method of claim 4 , wherein the corresponding forwarding class-weight is incremented.
6. The method of claim 4 , wherein the corresponding forwarding class-weight is incremented by 1.
7. The method of claim 4 , wherein the corresponding forwarding class-weight is decremented.
8. The method of claim 4 , wherein the corresponding forwarding class-weight is decremented by 1.
9. The method of claim 1 , if the error is higher than the previous iteration, reverting the forwarding class-weight to the previous value.
10. The method of claim 1 , further comprising identifying the forwarding class with the highest individual error and adapting the corresponding forwarding class-weight.
11. The method of claim 1 , wherein adapting the input distribution array involves decrementing a forwarding class-specific weight of the input distribution array.
12. The method of claim 1 , wherein adapting the input distribution array involves ordering a normalized traffic distribution, Dn, by individual error and adapting the forwarding class-weight with the greatest error.
13. The method of claim 1 , wherein adapting the input distribution array involves calculating an error as:
E
=
(
√
(
w
0
-
d
0
)
2
+
√
(
w
1
-
d
i
)
2
+
√
(
w
2
-
d
2
)
2
+
√
(
w
3
-
d
3
)
2
…
√
(
w
i
-
d
i
)
2
i
+
1
)
;
where w i is the forwarding class-specific normalized weight and d i is the forwarding class-specific normalized input traffic distribution, where i is an integer set of forwarding classes with i being an integer greater than 1.
14. The method of claim 13 , further comprising identifying the forwarding class with the highest individual error and adapting the corresponding forwarding class-weight.
15. The method of claim 14 , wherein the corresponding forwarding class-weight is incremented.
16. The method of claim 14 , wherein the corresponding forwarding class-weight is decremented.
17. A computing device for shaping traffic across a wide area network, the computing device comprising a memory and processor, the memory comprising instructions that, when executed by the processor, cause the processor to perform steps comprising:
advertising an input rate limit across a WAN from a first node;
advertising an input distribution array across the WAN from the first node, wherein the input distribution array includes forwarding class-specific weights; and
adapting at least one of the input rate limit and the input distribution array in response to an error between a target forwarding class distribution and an observed input traffic distribution;
wherein adapting the input distribution array involves calculating an error as:
E=RMSE(Wn, Dn),
where,
i=integer set of forwarding classes with i being an integer greater than 1,
a normalized array of weights, Wn=w 0 /max(W); . . . w i /max(W), where max(W) is the highest weight, w, amongst the weights of the set of forwarding classes, and
a normalized traffic distribution, Dn=r 0 /max(R), . . . r i /max(R), where max(R) is the highest rate, r, amongst the rates of the set of forwarding classes.
18. A computer readable storage medium that stores computer readable instructions, which when executed by a processor, cause the processor to perform steps comprising:
advertising an input rate limit across a WAN from a first node;
advertising an input distribution array across the WAN from the first node, wherein the input distribution array includes forwarding class-specific weights; and
adapting at least one of the input rate limit and the input distribution array in response to an error between a target forwarding class distribution and an observed input traffic distribution;
wherein adapting the input distribution array involves calculating an error as:
E=RMSE (Wn, Dn),
where,
i=integer set of forwarding classes with i being an integer greater than 1,
a normalized array of weights, Wn=w 0 /max(W); . . . w i /max(W), where max(W) is the highest weight, w, amongst the weights of the set of forwarding classes, and
a normalized traffic distribution, Dn=r 0 /max(R), . . . r i /max(R), where max(R) is the highest rate, r, amongst the rates of the set of forwarding classes.
19. A method for managing traffic across a wide area network (WAN), the method comprising:
advertising an input rate limit across a WAN from a first node;
advertising an input distribution array across the WAN from the first node, wherein the input distribution array includes forwarding class-specific weights; and
adapting at least one of the input rate limit and the input distribution array in response to an error between a target forwarding class distribution and an observed input traffic distribution;
wherein adapting the input distribution array involves ordering a normalized traffic distribution, Dn, by individual error and adapting the forwarding class-weight with the greatest error.
20. A method for managing traffic across a wide area network (WAN), the method comprising:
advertising an input rate limit across a WAN from a first node;
advertising an input distribution array across the WAN from the first node, wherein the input distribution array includes forwarding class-specific weights; and
adapting at least one of the input rate limit and the input distribution array in response to an error between a target forwarding class distribution and an observed input traffic distribution;
wherein adapting the input distribution array involves calculating an error as:
E
=
(
√
(
w
0
-
d
0
)
2
+
√
(
w
1
-
d
i
)
2
+
√
(
w
2
-
d
2
)
2
+
√
(
w
3
-
d
3
)
2
…
√
(
w
i
-
d
i
)
2
i
+
1
)
;
where w i is the forwarding class-specific normalized weight and d i is the forwarding class-specific normalized input traffic distribution, where i is an integer set of forwarding classes with i being an integer greater than 1.