draft zebra: add LSP entry to nexthop via recursive (part 2) #15331
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Commands added: r3# configure r3(config)# mpls fec MPLS FEC table label Label configuration ldp Label Distribution Protocol lsp Establish label switched path r3(config)# mpls fec mpls fec nexthop-resolution Authorise nexthop resolution over all labeled routes. r3(config)# mpls fec mpls fec nexthop-resolution r3# configure r3(config)# vrf default r3(config-vrf)# mpls fec MPLS FEC table r3(config-vrf)# mpls fec mpls fec nexthop-resolution Authorise nexthop resolution over all labeled routes. r3(config-vrf)# mpls fec nexthop-resolution east-vm# show running-config Building configuration... ... ! mpls fec nexthop-resolution ! ... Signed-off-by: Dmytro Shytyi <[email protected]>
Upon reconfiguring nexthop-resolution updates, update the lsp entries accordingly. If fec nexthop-resolution becomes true, then call again fec_change_update_lsp() for each fec entry available. If fec nexthop-resolution becomes false, then call again fec_change_update_lsp() for each fec entry available, and if the update fails, uninstall any lsp related with the fec entry. In the case lsp_install() and no lsp entry could be created or updated, then consider this call as a failure, and return -1. Co-developed-by: Dmytro Shytyi <[email protected]> Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>
This commit addresses an issue that happens when using bgp labeled unicast peering with a rr client, with a received prefix which is the local ip address of the bgp session. When using bgp ipv4 labeled session, the local prefix is received by a peer, and finds out that the proposed prefix and its next-hop are the same. To avoid a route loop locally, no nexthop entry is referenced for that prefix, and the route will not be selected. As it has been done for ipv4-unicast, apply the following fix for labeled address families: when the received peer is a route reflector, the prefix has to be selected, even if the route can not be installed locally. Fixes: f874552 ("bgpd: authorise to select bgp self peer prefix on rr case") Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>
There are two ways of iterating over nexthops of a given route entry. - Either only the main nexthop are taken into account (which is the case today when attempting to install an LSP entry on a BGP connected labeled route. - Or by taking into account nexthops that are resolved and linked in nexthop->resolved of the previous nexthop which has RECURSIVE flag set. This second case has to be taken into account in the case where recursive routes may be used to install an LSP entry. Introduce a new API in nexthop that will parse over the appropriate nexthop, if the nexthop-resolution flag is turned on or not on the given VRF. Use that API in the lsp_install() function so as to walk over the appropriate nexthops. Co-developed-by: Dmytro Shytyi <[email protected]> Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>
When an LSP entry is created from a FEC entry, multiple labels may now be appended to the LSP entry, instead of one single. Upon lsp creation, the LSP trace will display all the labels appended. Signed-off-by: Dmytro Shytyi <[email protected]>
Until now, when a FEC entry is added in zebra, driven by the reception of a BGP labeled unicast update, an LSP entry is created. That LSP entry is resolved by using the route entry which is also installed by BGP labeled unicast. This route entry is not available when we face with i-bgp peering session. I-BGP labeled sessions are used to establish IP connectivity across separate IGPs. The below dumps illustrate a 3 IGP topology. An attempt to create connectivity between the north and the south machines is done. The 3 separate IGPs are configured in Segment routing: - north-east - east-west - west-south We create BGP peerings between each endpoint of each IGP: - iBGP between (north) and (east) - iBGP between (east) and (west) - iBGP between (west) and (south) Before that patch, the FEC entries could not be resolved on the east machine: Before: east-vm# show mpls fec 192.0.2.1/32 Label: 18 Client list: bgp(fd 48) 192.0.2.5/32 Label: 17 Client list: bgp(fd 48) 192.0.2.7/32 Label: 19 Client list: bgp(fd 48) east-vm# show mpls table Inbound Label Type Nexthop Outbound Label -------------------------------------------------------- 1011 SR (OSPF) 192.168.2.2 1011 1022 SR (OSPF) 192.168.2.2 implicit-null 11044 SR (IS-IS) 192.168.3.4 implicit-null 11055 SR (IS-IS) 192.168.3.4 11055 30000 SR (OSPF) 192.168.2.2 implicit-null 30001 SR (OSPF) 192.168.2.2 implicit-null 36000 SR (IS-IS) 192.168.3.4 implicit-null east-vm# show ip route [..] B 192.0.2.1/32 [200/0] via 192.0.2.1 inactive, label implicit-null, weight 1, 00:17:45 O>* 192.0.2.1/32 [110/20] via 192.168.2.2, r3-eth0, label 1011, weight 1, 00:17:47 O>* 192.0.2.2/32 [110/10] via 192.168.2.2, r3-eth0, label implicit-null, weight 1, 00:17:47 O 192.0.2.3/32 [110/0] is directly connected, lo, weight 1, 00:17:57 C>* 192.0.2.3/32 is directly connected, lo, 00:18:03 I>* 192.0.2.4/32 [115/20] via 192.168.3.4, r3-eth1, label implicit-null, weight 1, 00:17:59 B 192.0.2.5/32 [200/0] via 192.0.2.5 inactive, label implicit-null, weight 1, 00:17:56 I>* 192.0.2.5/32 [115/30] via 192.168.3.4, r3-eth1, label 11055, weight 1, 00:17:58 B> 192.0.2.7/32 [200/0] via 192.0.2.5 (recursive), label 19, weight 1, 00:17:45 * via 192.168.3.4, r3-eth1, label 11055/19, weight 1, 00:17:45 [..] After command "mpls fec nexthop-resolution" was applied, the FEC entries will resolve over any non BGP route that has a labeled path selected. east-vm# show mpls table Inbound Label Type Nexthop Outbound Label -------------------------------------------------------- 17 SR (IS-IS) 192.168.3.4 11055 18 SR (OSPF) 192.168.2.2 1011 19 BGP 192.168.3.4 11055/19 1011 SR (OSPF) 192.168.2.2 1011 1022 SR (OSPF) 192.168.2.2 implicit-null 11044 SR (IS-IS) 192.168.3.4 implicit-null 11055 SR (IS-IS) 192.168.3.4 11055 30000 SR (OSPF) 192.168.2.2 implicit-null 30001 SR (OSPF) 192.168.2.2 implicit-null 36000 SR (IS-IS) 192.168.3.4 implicit-null Co-developed-by: Dmytro Shytyi <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]> Signed-off-by: Philippe Guibert <[email protected]>
Change the comment in the code that refers to ZEBRA_FLAG_XXX defines. Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>
Because the nhlfe label stack may contain more than one label, ensure to copy all labels. Co-developed-by: Dmytro Shytyi <[email protected]> Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>
There are 3 tests with OSPF, IS-IS, BGP and MPLS configured: 1. Check the status of BGP session between North and South == Established 2. Check the connectivity with "ping South -I North" 3. Check the label on the West. Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]> Signed-off-by: Louis Scalbert <[email protected]>
FEC nexthop entry resolution over MPLS networks Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>
|
closing duplicate #15259 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Duplicate of #15259
Need to test the CI after a rebase
Putting into draft