On Mon, Oct 22, 2018 at 01:17:30PM +0200, Marko Cupa? wrote:
> Hi,
>
> I am struggling to announce nexthop to my bgp peers after default
> ruleset change in 6.4's bgpd.conf.
>
> On 6.3, I used to have:
>
> match to $ISP1 set nexthop $CARP_TO_ISP1
> match to $ISP2 set nexthop $CARP_TO_ISP2
> deny from ebgp
> deny to ebgp
> allow to { $ISP1 $ISP2 }
> allow from ibgp
> allow to ibgp
> (...defaults...)
>
>
> I like the idea of having my simple ruleset done with minimal override
> to defaults. Moreover, I see that slapping above ruleset to 6.4 does
> not work the same as on 6.3 (I think I'm sending garbage upstream).
You can check with 'bgpctl show rib out nei $ISP1 detail' what you are
sending. Also tcpdump is able to show you what you are sending.
> Any good soul out there to tell me what to put above:
>
> ### for simple BGP setups, no editing below this line is required ###
>
> ...in order to set nexthop per upstream neighbor, if possible?
The new ruleset has a few deny quick rules in it. Make sure you don't hit
one of those.
It would be helpful to see the full ruleset as shown with 'bgpd -nv'
--
:wq Claudio
