On 4/14/16, 9:40 AM, "Acee Lindem (acee)" <[email protected]> wrote:

Acee:

...
>...I¹d leave the section 4 text as is
>and modify the section 3 text.

I agree with your proposal.

The reason for proposing a change in 4 is because if only one path exists
it may result in a loop (if pre-rfc1583 and current routers are present):

<-- 0/0 -- rtr1 -- rtr2 -- rtr3 -- destination -->

In this network there's a default route going left and the destination we
want to reach is on the right.  rtr2 is a pre-rfc1583 router and rtr3 uses
rfc6987.

If the source is connected to rtr1, rtr1 will decide to use rtr3 as
transit (it is the only path).  But rtr2 will not and send the packet back
to rtr1 (following the default), resulting in a loop.

...so the text below clarifies that it is a possibility.


We can file these are independent errata and let the AD deal with them.
;-)  In both cases I think the update doesn't change the spirit of the RFC
and can just be help for document update.

Thanks!

Alvaro.


>>
>>NEW>
>>4.  Deployment Considerations
>>...
>>   on using them rather than the path through the stub router.  If the
>>   path through the stub router is the only one, the routers of the
>>   first type will not use the stub router for transit, while the routers
>>   of the second type will still use this path, which may result in a
>>routing 
>>   loop.
>>...
>>

_______________________________________________
OSPF mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ospf

Reply via email to