Objectives
The simulator ticks each one as soon as the network meets it.
On PE1 create segment list BOTTOM: labels 16003 (P2), 16004 (P3), 16005 (PE2).
Create policy BLUE, color 20, end-point 5.5.5.5, with candidate preference 100 explicit segment-list BOTTOM. It comes up through P2 and P3.
Add autoroute include all: PE1’s IS-IS route to 5.5.5.5 now uses policy BLUE.
PC1 → PC2 (LAN 10, no color) now rides policy BLUE through P2 and P3.
Before you start
autoroute steers by destination, not by color: every IGP route whose path runs through the end-point, and every BGP route whose next hop is behind it, uses the policy. The color still names the policy, but no route needs to carry it.
Concept
SR-TE policies and candidate paths
An SR-TE policy is identified by a color and an end-point. It carries a list of candidate paths with preferences. A dynamic candidate is computed on the head end over the TE database. An explicit candidate is a segment list of prefix-SIDs and adjacency-SIDs. The valid candidate with the highest preference is active.
Concept
Steering traffic into a policy: color and autoroute
Routes are steered into an SR-TE policy in two ways. With automated steering, a BGP route whose color extended community equals the policy’s color, and whose next hop equals the policy’s end-point, uses the policy. With autoroute, every IGP route through the end-point uses it, color or not. Everything else takes the default path: the SR (or LDP) shortest path.
