BGP route remains inactive in RIB and not selected as FIB route

[deepakkumarbhagat]

FRR version: 8.2.2

Setup

R1(20:1::1/64)--------------------directly connected network---------------------(20:1::2/64)R2----------------30:1::/64
(loopback-11:11:11::11/128) ----------------------------------------------------------------------------(loopback-22:22:22::22/128)

Observation,

I have a setup with two routers as bgp peers shown above, the route advertised by a peer R2 is not installed in FIB and remains inactive in RIB at R1, gonig by the existing frr issues/discussions, my understanding is zebra doesn't install the route to FIB if the next_hop in the route is not reachable, is this correct ?? The next_hop in my case is R2 loopback interface and we do have a static route added for this at R1. But I do see this static route is removed/added randomly not very sure who is triggering this static route removal/addition, but I suspect the situation I'm hiting is when the route is being received by the router , at the same time this static route removal/addition is happening and this may be leading to stuck route as inactive in RIB.

But my expectation here is, this route should be selected as FIB route later when the next_hop becomes reachable again ., is this expectation correct or the behaviour of keeping the route as inactive even if next_hop becomes reachable later is correct??

R1 RIB
VRF default:
5514 C>* 10:1::/64 is directly connected, net691c27d41820, 00:01:53
5515 C>* 11:11:11::11/128 is directly connected, loopback0, 00:01:54
5516 C>* 20:1::/64 is directly connected, net124dce659ffb, 00:01:52
5517 S> 22:22:22::22/128 [1/0] via 20:1::2, net124dce659ffb, weight 1, 00:00:03 --> this is next_hop reachability (the entry age is only 3 secs old)*
5518 B 30:1::/64 [20/0] via 22:22:22::22 inactive, weight 1, 00:01:06 -------------> this remains inactive (this entry is 1 mins 6 secs old)
5519 C * fe80::/64 is directly connected, loopback0, 00:01:54
5520 C * fe80::/64 is directly connected, net124dce659ffb, 00:01:57
5521 C>* fe80::/64 is directly connected, net691c27d41820, 00:02:00

show bgp vrf all
5716
5717 Instance default:
5718 BGP table version is 8, local router ID is 0.0.0.1, vrf id 0
5719 Default local pref 100, local AS 10
5720 Status codes: s suppressed, d damped, h history, * valid, > best, = multipath,
5721 i internal, r RIB-failure, S Stale, R Removed
5722 Nexthop codes: @NNN nexthop's vrf id, < announce-nh-self
5723 Origin codes: i - IGP, e - EGP, ? - incomplete
5724 RPKI validation codes: V valid, I invalid, N Not found
5725
5726 Network Next Hop Metric LocPrf Weight Path
5727 *> 10:1::/64 :: 0 32768 i
5728 *> 30:1::/64 22:22:22::22 0 0 20 i

[deepakkumarbhagat]

@ton31337 , @riw777
can you please clarify if the behaviour is expected or it's a bug ??

[deepakkumarbhagat]

awaiting comments @ton31337

[ton31337]

Enable debug bgp nht, show also show bgp nht.

[deepakkumarbhagat]

Thanks @ton31337 for your response, the issue is not seen as of now, I shall be able to share the logs as soon as the issue is seen again.