[j-nsp] AS path preservation when importing from instance.inet.0 to inet.0

[Dragan Jovicic]

Awesome ! :)

Best
Dragan

On Mon, Oct 17, 2016 at 5:43 PM, Paul S. <contact at winterei.se> wrote:

> Hi Dragan,
>
> That worked, actually. Many thanks!
>
> As to the lt interface, will read up as suggested.
>
>
> On 10/17/2016 03:18 AM, Dragan Jovicic wrote:
>
> You could try import-rib policy, where you match on bgp routes and as-path
> prepend, something like this:
> set routing-options rib-groups rib-1 import-policy pl-1
> set policy-options policy-statement pl-1 term 10 from proto bgp
> set policy-options policy-statement pl-1 term 10 then as-path-prepend
> "XXXX"
> set policy-options policy-statement pl-1 term 10 then accept
>
> Try and let me know please.
>
> [quote]
> Speaking of the lt tunnel, is there any clear drawbacks to using it? Once
> upon a time, I recall hearing that it was bandwidth constrained. I'm doing
> this on a Trio MX.
> [/quote]
>
> There is a bandwidth limitation, check out docs. As for when to use it,
> depends...
>
> Best
>
> Dragan
>
>
>
> On Sun, Oct 16, 2016 at 6:05 PM, Paul S. <contact at winterei.se> wrote:
>
>> Hi guys,
>>
>> So, in a bit of a peculiar situation. I think rather than explaining it,
>> it's possibly easier to express through configs. I've added it at the end
>> of the email.
>>
>> Basically, my local-as in a ri is different compared to my local-as set
>> in the master instance. When I import a BGP route (that I'm actually
>> originating in the RI and would like to originate in the master instance
>> too), and then export it to other peers -- the originating ASN gets
>> rewritten to the master instance's ASN instead.
>>
>> For example - AS-path in RI A for 20.20.20.0/24 is [2] I
>>
>> When imported via instance-import to inet.0 and exported to other peers,
>> I can see that the AS-path becomes [1] I. What I'd like it to be is [1 2]
>> I, i.e: the RI looks like a downstream adjacency of the master instance
>> instead.
>>
>> Is there any way to achieve this (other than setting up a lt tunnel and
>> peering with the master)? Speaking of the lt tunnel, is there any clear
>> drawbacks to using it? Once upon a time, I recall hearing that it was
>> bandwidth constrained. I'm doing this on a Trio MX.
>>
>> Pointers welcome, thank you for reading.
>>
>> (As to why the multi-asn stupidity, that's due to a limitation on our
>> upstream provider's side. Sadly, no control over that)
>>
>>
>> Config from the "*master*" instance:
>>
>>
>> routing-options {
>>
>>     router-id 1.1.1.1;
>>
>>    autonomous-system 1;
>>
>> }
>>
>> protocols
>>
>> {
>>
>>     bgp {
>>
>>        nei 10.10.10.10 peer-as 500;
>>
>>    }
>>
>> }
>>
>>
>> Config from a *second* routing-instance
>>
>> A {
>>
>>     instance-type virtual-router;
>>
>>     interface x;
>>
>>     routing-options {
>>
>>         router-id 2.2.2.2;
>>
>>     }
>>
>>     protocols { bgp {
>>
>>         nei 10.10.10.15 peer-as 500;
>>
>>         nei 10.10.10.15 local-as *2;*
>>
>>      }
>>
>> }
>>
>> _______________________________________________
>> juniper-nsp mailing list juniper-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/juniper-nsp
>>
>
>
>