|
|
1 |
(% class="row" %) |
|
|
2 |
((( |
|
|
3 |
(% class="col-xs-12 col-sm-7 test-report-content" %) |
|
|
4 |
((( |
|
|
5 |
---- |
|
|
6 |
|
|
|
7 |
As indicated in RFC 9256, there are several approaches that a headend router may use to take a given flow and place it into an SR policy, one of them being per‐destination steering. This architecture allows service operators to create separate forwarding paths, commonly known as candidate paths, and optimally distribute the traffic across several network links or even prioritize some paths for high‐priority services. We deployed a spine‐and‐leaf topology with two spine routers and several PEs to perform the test. As illustrated in the figure below, it allowed two paths to each PE for IPv4 and IPv6 traffic. |
|
|
8 |
|
|
|
9 |
[[Figure 82: SR-MPLS SRTE Traffic Steering>>image:433705802207985665_SRMPLS-10-SRTE-Generic-5-v1.png||alt="Figure 82" width="550"]] |
|
|
10 |
|
|
|
11 |
From each edge, we advertised a "color" community from each PE router to the destination router over BGP, where an SR policy could match the endpoint's color and address to generate an appropriate SID list. |
|
|
12 |
With the application of the color tag, the corresponding SR policy is implemented in the network, directing packets through a specific designated SR-TE path. Upon removing the color tag, the traffic switched to an alternate SR path; if no SR path was available, the traffic was dropped as expected. We tested all source-destination pairs in the network and confirmed that no packet loss occurred in any of the SR-TE tunnels. |
|
|
13 |
In general, the results presented in this section demonstrate the effectiveness of per-destination SR-TE forwarding in SR-MPLS environments. This advancement provides operators with tools to optimize routing schemes, balance traffic loads, and ensure high availability for critical services. |
|
|
14 |
|
|
|
15 |
{{container cssClass="tc-role-table"}} |
|
|
16 |
(% class="table-bordered" %) |
|
|
17 |
|=PE|=Spine|=Traffic Generator |
|
|
18 |
|((( |
|
|
19 |
Arista 7280R3, |
|
|
20 |
Arrcus S9610-36D, |
|
|
21 |
Ciena 5134, |
|
|
22 |
Ciena 8140 Coherent Metro Router, |
|
|
23 |
H3C CR16000-M1A, |
|
|
24 |
Huawei NetEngine 8000 M14, |
|
|
25 |
Huawei NetEngine A816, |
|
|
26 |
Juniper ACX7024, |
|
|
27 |
Nokia 7730 SXR-1x-44s, |
|
|
28 |
Nokia 7750 SR-1, |
|
|
29 |
Ribbon NPT-2100 |
|
|
30 |
)))|((( |
|
|
31 |
Arista 7280R3, |
|
|
32 |
Juniper ACX7100-48L |
|
|
33 |
)))|Keysight IxNetwork |
|
|
34 |
|
|
|
35 |
Table 25: SR-MPLS - SRTE Traffic Steering (per destination) - per destination |
|
|
36 |
{{/container}} |
|
|
37 |
(% id="prev-next-links" %) |
|
|
38 |
|[[< Previous>>doc:SR-MPLS Inter AS]]|[[Next ~>>>doc:Liveness of SR Policy using Seamless BFD]] |
|
|
39 |
))) |
|
|
40 |
|
|
|
41 |
(% class="col-xs-12 col-sm-5 test-report-sidebar" %) |
|
|
42 |
((( |
|
|
43 |
{{box}} |
|
|
44 |
{{include reference="Main.Multi-Vendor MPLS & SDN Interoperability Test Report 2025.Sidebar Nav"/}} |
|
|
45 |
{{/box}} |
|
|
46 |
))) |
|
|
47 |
))) |