Bill,
I've had a symptom a bit like this when I've done something foolish like
router bgp 1234
network 192.10.10.0
ip route 192.10.10.0 255.255.255.0 null0 254
^^^
(oops)
When another route gets into the FIB with a lower administrative weight
you see a flap.
Not sure how you can get a lower admin weight than connected though.
Is the interface line protocol flapping?
I've also seen the same happen on larger blocks using an aggregate address,
something smaller than a class C and accidentally not having
auto-summary switched off on another router. But that's not your
problem here.
How are you originating the route in BGP? Redist connected? network?
Try debugging the BGP session and see if you see withdrawals and
announcements.
Alex Bligh
GX Networks (formerly Xara Networks)
>
> Ok, I'm at ropes end....
>
> >sh ip route 198.32.64.12
> Routing entry for 198.32.64.0/24
> Known via "connected", distance 0, metric 0 (connected)
> Redistributing via bgp 4555
> Advertised by bgp 4555 metric 50 weight 3500
> Routing Descriptor Blocks:
> * directly connected, via FastEthernet2/0
> Route metric is 0, traffic share count is 1
> >sh ip bgp 198.32.64.12
> BGP routing table entry for 198.32.64.0/24, version 1618391
> Paths: (1 available, best #1, advertised over EBGP)
> Local
> 0.0.0.0
> Origin IGP, localpref 100, weight 32768, valid, sourced, local, best
> >sh ip bgp sum
> ...
> 198.32.146.10 4 226 785093 20878 1617928 0 0 1w4d 51493
> ...
>
> from 198.32.146.10
> >sh ip route 198.32.64.12
> % Network not in table
> >sh ip bgp 198.32.64.12
> BGP routing table entry for 198.32.64.0/24, version 14990982
> Paths: (1 available, no best path, advertised over IBGP, EBGP)
> 4555 (history entry)
> 198.32.146.11 from 198.32.146.11 (192.0.2.10)
> Origin IGP, metric 1, external
> Dampinfo: penalty 11024, flapped 331 times in 04:21:30
> >
>
> Any pointers on why the cisco code might think this route is "flapping"?
>
> --bill
This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:13:13 EDT