sonic-buildimage/dockers/docker-fpm-frr/frr
pavel-shirshov 84405ab953 [bgp]: Enable next-hop-tracking through default (#5600)
**- Why I did it**
FRR introduced [next hop tracking](http://docs.frrouting.org/projects/dev-guide/en/latest/next-hop-tracking.html) functionality.
That functionality requires resolving BGP neighbors before setting BGP connection (or explicit ebgp-multihop command). Sometimes (BGP MONITORS) our neighbors are not directly connected and sessions are IBGP. In this case current configuration prevents FRR to establish BGP connections.  Reason would be "waiting for NHT". To fix that we need either add static routes for each not-directly connected ibgp neighbor, or enable command `ip nht resolve-via-default`

**- How I did it**
Put `ip nht resolve-via-default` into the config

**- How to verify it**
Build an image. Enable BGP_MONITOR entry and check that entry is Established or Connecting in FRR

Co-authored-by: Pavel Shirshov <pavel.contrib@gmail.com>

Signed-off-by: Abhishek Dosi <abdosi@microsoft.com>
2020-10-13 22:42:29 -07:00
..
bgpd [bgp] Add 'allow list' manager feature (#5513) 2020-10-06 11:15:19 -07:00
common Skip vnet-vxlan interfaces from generating networks (#5251) 2020-09-03 15:49:59 -07:00
staticd [bgpcfgd]: Split one bgp mega-template to chunks. (#4143) 2020-04-25 09:41:28 +00:00
zebra [bgp]: Enable next-hop-tracking through default (#5600) 2020-10-13 22:42:29 -07:00
frr.conf.j2 [bgpcfgd]: Split one bgp mega-template to chunks. (#4143) 2020-04-25 09:41:28 +00:00
isolate.j2 [bgpcfgd]: Split one bgp mega-template to chunks. (#4143) 2020-04-25 09:41:28 +00:00
unisolate.j2 [bgpcfgd]: Split one bgp mega-template to chunks. (#4143) 2020-04-25 09:41:28 +00:00