Enable TI-LFA Fast Reroute for ISIS-SR¶

Preparing The Lab¶

Log into the LabAccess jumpserver:
1. Type labs, or select Option 97 to get to the Additional Labs menu.
2. Type or select the option for mesh-topology-evpn-labs in order to get to the EVPN labs.
3. Type tilfa in this menu to configure the topology with the necessary prerequisites.

Lab Tasks¶

Enable Topology Independent Loop Free Alternate, or TI-LFA, calculation to occur on EOS1.
1. First, enable TI-LFA to protect against link failures in the Level-2 IS-IS topology.
  
  Note
  
  TI-LFA will calculate backup paths between the IS-IS enabled nodes assuming that the primary best path has failed. In the event of an actual failure, hardware forwarding would switch to this backup path in 50 ms. This would normally be paired with BFD monitoring. In the virtual ATD labs, it is not possible to simulate actual failures, but the TI-LFA control-plane can still be validated.
```
router isis 100
    !
    address-family ipv4 unicast
        fast-reroute ti-lfa mode link-protection level-2
```
2. Set a delay so that traffic will transition to the post-failure path only after the network has fully converged.
  
  Note
  
  Normally, IS-IS would use the TI-LFA calculated path only until the local router has reconverged after the failure. To prevent micro-loops during failure events, we introduce a delay for the router to switch from the TI-LFA calculated backup path to give the network time to converge globally.
```
router isis 100
    timers local-convergence-delay protected-prefixes
```

Repeat the above configuration steps on all other Service Provider nodes.

Configure EOS2 - EOS8 for TI-LFA calculation.

router isis 100
    timers local-convergence-delay protected-prefixes
    !
    address-family ipv4 unicast
        fast-reroute ti-lfa mode link-protection level-2

Testing¶

Verify local ISIS-SR TI-LFA status and forwarding on EOS1.

Display the Node SIDs of the topology and observe that some now have Protection.

Note

You will notice some prefix-segments are not protected. This is due to that fact that an ECMP route for those Nodes is already present, so there is no need to further calculate a backup path. If a link fails in the ECMP group, it will automatically use the other. Once a single link is the best path, it would then calculate a TI-LFA backup path. You can observe which nodes have an ECMP route with show ip route and verifying which prefixes have more than one next-hop listed.
```
show isis segment-routing prefix-segments
```
Display the logic EOS1 uses to calculate the backup path to the other nodes in the topology.

Abstract

The Constraint is how EOS displays what would happen in the event of a given failure; basically following the format “If Constraint is true, then use Path to reach Destination.”
```
show isis ti-lfa path
```
Display primary and backup path information from EOS1 to EOS8.

Note

We can check how EOS1 will reach EOS8 by first looking up the SR tunnel for the EOS8 node prefix 8.8.8.8/32. Then we can check the TI-LFA tunnel index, which in the below example happened to be 7 though this may vary in your lab. Lastly, we can verify that the MPLS hardware table has programmed the label corresponding to the EOS8 Node-SID to use the TI-LFA Tunnel.
```
show isis segment-routing tunnel 8.8.8.8/32
show isis ti-lfa tunnel 7
show mpls lfib route 900008
```
Verify L2VPN routes towards EOS7 are using the TI-LFA tunnel from EOS3.

Note

We will trace the MAC of EOS10, which in this example is 1426.0c23.74e4. You should replace this in the commands below with the MAC of Et1 on EOS10 which can be found in the command show interface Ethernet1. Likewise the tunnel index of 3 should be replaced with the index found in parenthesis from the l2rib output.
```
show l2rib output mac 1426.0c23.74e4
show tunnel fib isis segment-routing 3
```

Success

Lab Complete!