exabgp 4.0.6: Wrong default route decoding

[yannick thomas]

exabgp 4.0.6 causes wrong decoding when receiving route announced/withdrawn for default prefix (0.0.0.0/0)

gobgp
-----

10:07:58.492463 IP (tos 0x0, ttl 255, id 39934, offset 0, flags [DF], proto TCP (6), length 162)
    192.168.80.41.bgp > 192.168.80.38.39290: Flags [P.], cksum 0x2235 (incorrect -> 0xda7e), seq 5777:5887, ack 267, win 227, options [nop,nop,TS val 150076867 ecr 166857791], length 110: BGP
        Update Message (2), length: 110
          Origin (1), length: 1, Flags [T]: IGP
          AS Path (2), length: 0, Flags [T]: empty
          Local Preference (5), length: 4, Flags [T]: 100
          Originator ID (9), length: 4, Flags [O]: 192.168.80.37
          Cluster List (10), length: 4, Flags [O]: 1.2.3.4
          Multi-Protocol Reach NLRI (14), length: 29, Flags [O]:
            AFI: IPv4 (1), SAFI: labeled VPN Unicast (128)
            nexthop: RD: 0:0 (= 0.0.0.0), 192.168.80.37, nh-length: 12, no SNPA
              RD: 192.168.80.37:20, 0.0.0.0/0, label:31 (bottom)
          Extended Community (16), length: 24, Flags [OT]:
            unknown extd community typecode (0x030c), Flags [none]
              0x0000: 030c 0000 0000 000a
            unknown extd community typecode (0x0314), Flags [none]
              0x0000: 0314 0000 0000 0000
            target (0x0002), Flags [none]: 2:5005 (= 0.0.19.141)

bagpipe-bgp
-----------

juin 29 10:09:14 l-n4c-oscomp2 bagpipe-bgp[20401]: DEBUG networking_bagpipe.bagpipe_bgp.engine.exabgp_peer_worker.192-168-80-41 [-] protocol read message: 37.0.20.0/24 next-hop 192.168.80.37 rd 241.0.1.192:43088 origin igp local-preference 100 originator-id 192.168.80.37 cluster-list [ 1.2.3.4 ] extended-community [ target:2:5005 encap:MPLS consistentHashSortOrder:0 ] {{(pid=20401) _receive_loop_fun /opt/stack/networking-bagpipe/networking_bagpipe/bagpipe_bgp/engine/exabgp_peer_worker.py:295}}

[Thomas Morin (tmmorin-orange)]

I pushed a fix in exabgp (https://github.com/Exa-Networks/exabgp/pull/835), but we will have to ban exabgp 4.0.6 from the versions accepted for openstack.

[Thomas Morin (tmmorin-orange)]

For the record, the issue was introduced between 4.0.5 and 4.0.6 in
https://github.com/Exa-Networks/exabgp/commit/682947ff6356559a81d014a42eb3723b5ff75264 .

[Thomas Morin (tmmorin-orange)]

Fixed by https://review.openstack.org/#/c/579483/

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/requirements rocky-eol release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/requirements stein-eol release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/requirements yoga-eom release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/requirements train-eol release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/requirements ussuri-eol release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/requirements victoria-eom release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/requirements wallaby-eom release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/requirements xena-eom release.

[OpenStack Infra (hudson-openstack)]

This issue was fixed in the openstack/requirements zed-eom release.