Skip to Content.

rare-dev - Re: [rare-dev] Routing Internet by tunnel configuration

Subject: Rare project developers

List archive


Re: [rare-dev] Routing Internet by tunnel configuration


Chronological Thread 
  • From: Gabriel Tetzner <>
  • To:
  • Cc: , Frédéric LOUI <>
  • Subject: Re: [rare-dev] Routing Internet by tunnel configuration
  • Date: Fri, 18 Nov 2022 16:47:10 -0300

Hello Frédéric,

Eu acho que consegui fazer a vrf reconhecer a saída gateway

Fiz com hairpin11 e hairpin12 e acertei uma rota da vrf VXLAnN para 1.1.1.1 que seria minha saída gateway:



Em ter., 15 de nov. de 2022 às 18:01, Frédéric LOUI <> escreveu:
Hi,

What you are referring is called inter-VRF routing.
You have multiple ways to achieve that.

For example you can use:

1- multiple MPLS route-target  and from a site connected to a PE that import both RT, you can do the inter-VRF routing.
2- freeRtr has a nice feature feature in order to bon VRF locally that is called hairpin.  Juniper calls this feature logical-tunnel interface)
3- Other method, such as route redistribution between VRF etc.

-2- is good if you are not familiar with service provider MPLS CORE. (I’m not sure about your design choice)

Take a look at this test case:
http://docs.freertr.org/guides/reference/md/conn-hairpin01.tst/

It binds v1 to hairpin11 and v2 to hairpin12, routing is done through this interface.

The cool thing is that you can apply ACL on hairpin11 and hairpin12 in order to filter traffic.

Not sure if I understood and answered your question though :-/

I’ll let mc36 elaborate more if need.

All the best
Frederic

> Le 15 nov. 2022 à 17:30, Gabriel Tetzner <> a écrit :
>
> Hi all,
>
> sorry for the inconvenience and I know that everyone must be busy,
>
> I managed to make a wireguard client recognize our vxlan's tunnel infrastructure and reach another router from another available network, but I need to advance in one thing...
>
> Everything I am doing is separated into a vrf called VXLAN, but there is a global vrf called INET that freeRouter creates when it is installed in the virtual machine.
>
> This my global INET vrf has output to the Internet via an Ethernet1 interface that points to an outbound gateway, now consider that I want my other VXLAN vrf to have output to the Internet so that I have in my routes the external connectivity.
>
> I thought about creating a second eth2 interface and linking it with pcapInt afterwards, I haven't done that yet as I tried using a tapInt by default and within my research, I saw that you have a way to import the routes from one vrf to the other is what they kind of call leaking, not sure.
>
> Well, my question is: how can I make this vrf VXLAN communicate with the output (gateway), using some means that is possible for this?
>
> And on that note...
> The context of this ẽ́ that I am looking for isolation of my network and for it to point to the Internet afterwards, I need this to do tests with SD-WAN as well, since I believe vpdn needs an output to reach the destination target of the sd-wan server.
>
> Here are some configurations:
>
> prefix-list vrf_inet
>  sequence 10 permit 0.0.0.0/0 ge 0 le 0
>  exit             
> !                 
> route-map VRF_TO_NET
>  sequence 10 action permit
>  sequence 10 match prefix-list vrf_inet
>  !                 
>  exit             
> !                 
> route-map VRFINET_TO_VXLAN
>  sequence 10 action permit
>  sequence 10 match prefix-list vrf_inet
>  !                 
>  exit             
> !         
>  vrf definition inet
>  rd 200:200       
>  import4map VRF_TO_NET
>  export4map VRFINET_TO_VXLAN
>  exit             
> !                             
> vrf definition vxlan
>  description vxlan vrf isolation
>  rd 100:100       
>  import4map VRFINET_TO_VXLAN
>  export4map VRF_TO_NET
>  exit             
> !
> interface ethernet1
>  description uplink     
>  vrf forwarding inet
>  ipv4 address dynamic dynamic
>  ipv4 gateway-prefix all4
>  ipv4 dhcp-client enable
>  ipv4 dhcp-client early
>  ipv6 address dynamic dynamic
>  ipv6 gateway-prefix all6
>  ipv6 slaac-client enable
>  ipv6 prefix-suppress
>  router eigrp4 2 enable
>  no shutdown       
>  no log-link-change
>  exit             
> !
> vcpe1#show ipv4 route inet                                                     
> typ  prefix             metric  iface          hop            time
> DEF  0.0.0.0/0          0/2     ethernet1      200.x.x.65  00:00:21
> C    10.255.255.0/24    0/0     ethernet20001  null           12:05:01
> LOC  10.255.255.254/32  0/1     ethernet20001  null           12:05:01
> C    200.x.x.64/28   0/0     ethernet1      null           00:00:21
> REM  200.x.x.65/32   0/0     ethernet1      200.x.x.65  00:00:21
> LOC  200.x.x.69/32   0/1     ethernet1      null           00:00:21

> vcpe1#show ipv4 route vxlan                                                   
> typ  prefix            metric  iface      hop   time
> C    10.0.0.0/24       0/0     tunnel2    null  12:09:08
> LOC  10.0.0.2/32       0/1     tunnel2    null  12:09:08
> C    124.0.2.0/28      0/0     tunnel1    null  12:09:08
> LOC  124.0.2.2/32      0/1     tunnel1    null  12:09:08
> C    200.x.x.85/32  0/0     ethernet2  null  11:27:19
>
> This ethernet2 is the interface I tried to create ;)
>
> Best regards,Gabriel Tetzner Menegueti.
>




Archive powered by MHonArc 2.6.19.

Top of Page