RARE user and assistance email list

[Kaliraj Vairavakkalai <>]

>r3-r4 will have two rsvp-te tunels and vpnv4/vpnv6 ibgp,
>r4-r5 will have a single bgp-ct

 

Why rsvp-te tunnel end at r4? Are there two more rsvp-te tunnels from r4-r5?

And is there a bgp-ct session between r3-r4 also? With r4 doing nexthop-self?

 

Thanks

Kaliraj

From: mc36 <>
Date: Tuesday, February 14, 2023 at 10:49 PM
To: Kaliraj Vairavakkalai <>, <>, Krzysztof Szarkowicz <>, Simon Leinen <>, Ivana Golub <>, Frédéric LOUI <>
Cc: Natrajan Venkataraman <>, Reshma Das <>, <>
Subject: Re: [freertr] [RARE-users] EANTC - BGP CT - Juniper-FreeRTR

[External Email. Be cautious of content]


hi,
let's have a different topology to answer your question:

  r1                 r6
    > r3---r4---r5 <
  r2                 r7

r1 and r2 are cpes in vpnv4/vpnv6 vrfs of r3,
r3-r4 will have two rsvp-te tunels and vpnv4/vpnv6 ibgp,
r4-r5 will have a single bgp-ct
and r6 and r7 are cpes in vpnv4/vpnv6 vrfs of v5

so this doable, but i don't have a complete test case for it, just in pieces,
but i can quickly write you this topology as a test if that helps?
br,
cs



On 2/15/23 07:42, Kaliraj Vairavakkalai wrote:
>>rare/freertr used the neighbor 1.1.1.1 address-family knob whta controls, but the safis are similarly exclusive, only one of them is accepted:
>
> Yea, we noticed that during last hackathon. that   s ok, we can work with that.
>
> But my main question was this:
>
> Lets say there are two RSVP-TE-Tunnels (red, blue) between r1 to r2.
>
>>   red                 red---r5
>> r0-->r1====r2<--r4
>> blue                 blue
>
>> so in the above topology we can separate the two colors while using a single link in between r1 and r2 with 1 bgp-ct session...
>
> Can red bgp-ct routes put traffic on red-Tunnel. And blue-routes put traffic on blue-Tunnel?
>
> i.e. based on the mapping-community on the bgp-route, resolve bgp nexthop r2 over a red/blue-Tunnel.
>
> I thought that is the    Colorful resolution    functionality that arrived in spring. Sorry if I misunderstood. :)
>
> Thanks
>
> Kaliraj
>
> *From: *mc36 <>
> *Date: *Tuesday, February 14, 2023 at 10:11 PM
> *To: *Kaliraj Vairavakkalai <>, Krzysztof Szarkowicz <>, Simon Leinen <>, Ivana Golub
> <>, Fr  d  ric LOUI <>
> *Cc: *Natrajan Venkataraman <>, Reshma Das <>, <>
> *Subject: *Re: [freertr] [RARE-users] EANTC - BGP CT - Juniper-FreeRTR
>
> [External Email. Be cautious of content]
>
>
>> (xr lifted this recently, whereas it's still true for xe and rare/freerouter)
>>
>
>
> c8000v-1(config)#router bgp 1
> c8000v-1(config-router)#address-family ipv4 unicast
> c8000v-1(config-router-af)#neighbor 1.1.1.1 send-label ?
>       explicit-null   Advertise Explicit Null label in place of Implicit Null
>       <cr>                     <cr>
>
> c80000v-1(config-router-af)#neighbor 1.1.1.1 send-label
>
> so this knob is what "converts" this peering from ipv4 unicast to ipv4 labeled unicast, and obviously it's set or not... :)
>
> rare/freertr used the neighbor 1.1.1.1 address-family knob whta controls, but the safis are similarly exclusive, only one of them is accepted:
>
> sid#configure
> sid(cfg)#router bgp4 123 vrf v3
> sid(cfg-rtr)#neighbor 1.1.1.1 remote-as 321
> sid(cfg-rtr)#neighbor 1.1.1.1 address-family unicast labeled
> sid(cfg-rtr)#show running-config this
> router bgp4 123 vrf v3
>     neighbor 1.1.1.1 remote-as 321
>     neighbor 1.1.1.1 address-family labeled
>     exit
> !
>
> sid(cfg-rtr)#
>
>
> Juniper Business Use Only
>

Juniper Business Use Only