Hi,

If 6rd-A and 6rd-B belong to the same 6rd domain, I don't see the routing loops.

washam

----- Original Message -----
From: Rémi Després <[email protected]>
Date: Thursday, January 21, 2010 4:04 pm
Subject: Re: [Softwires] routing loop issue//re:  Please review 6rd
To: Xu Xiaohu <[email protected]>
Cc: [email protected]


>  Le 21 janv. 2010 à 08:42, Xu Xiaohu a écrit :
>  
>  
>  > 
>  >    +------------------------+
>  >    |                        |
>  >    |                        |
>  >    |     IPv6 Internet      |
>  >    |                        |
>  >    +------+----------+------+
>  >           |          |
>  >           |          |
>  >        +--+--+     +--+--+
>  >    +---+6rd-A+-----+6rd-B+--+
>  >    |   +-----+     +-----+  |
>  >    |                        |
>  >    |        IPv4 ISP        |
>  >    |       +-------+        |
>  >    +-------+ CPE-1 +--------+
>  >            +---+---+
>  >                |
>  >          +-----+-----+
>  >          |           |
>  >          | IPv6 Site |
>  >          +-----------+
>  > 
>  > In the above scenario, two 6rd BRs are configured with the same 
> address for
>  > anycast service. Assuming the nearest 6rd BR for the CPE is 6rd-A, 
> all IPv6
>  > packets from the IPv6 site and destined to the IPv6 internet will be
>  > tunneled towards 6rd-A by the CPE. Once 6rd-A lost connections to 
> the IPv6
>  > Internet, some failover mechanism should be used to direct these packets
>  > towards 6rd-B.
>  
>  > One possible solution is: 6rd-A withdraws the route to the
>  > IPv6 Internet, which was advertised in IPv4 ISP network previously, 
> as soon
>  > as its connection to the IPv6 Internet is lost.
>  
>  This is in my understanding the right solution, precisely because 
> protection against routing loops can be achieved by preventing tunnels 
> between BRs.
>  
>  
>  > Another solution is:
>  > establish a softwire tunnel between the two 6rd BRs for failover 
> route. With
>  > the latter solution, some normal packets tunneled in the softwire during
>  > failover will be filtered mistakenly.
>  
>  That's why the other solution should be preferred.
>  
>  RD
>  
>  _______________________________________________
>  Softwires mailing list
>  [email protected]
>  https://www.ietf.org/mailman/listinfo/softwires
>  
_______________________________________________
Softwires mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/softwires

Reply via email to