Thanks Bryan, I'm trying it in GNS right now.

On Wed, Aug 19, 2009 at 9:50 PM, Bryan Bartik <[email protected]> wrote:

> Vikas,
>
> Not sure if this will work in a non-mpls vpn environment, but cost
> community can be used to prefer the cloud over the backdoor link. Read more
> here:
>
>
> http://www.cisco.com/en/US/docs/ios/12_0s/feature/guide/s_bgpcc.html#wp1067175
>
> Would be interesting to lab...
>
>
> On Tue, Aug 18, 2009 at 6:52 PM, Vikas Sharma <[email protected]> wrote:
>
>> Hi Adrian,
>>
>> Answers to your questions.
>>
>> 1. No, ISP is not running MPLS within this cloud.
>>
>> 2. No, it's just one ISP.
>>
>> 3. We have 2 sites with high speed dark fibre connectivity between them
>> and this is the best path. There are multiple redundant paths between the
>> core switches at each site. The ISP has its routers at each location that
>> connect to the core switch.
>>
>> 4. No we won't be running BGP within the sites, only EIGRP.
>>
>> At each site are subnets, no 2 subnets span the other site so they are
>> exclusive to that site only.
>>
>> The route map is configured to look at the access list for site A (HQ).
>> This access list contains a list of networks within Site A. For routes that
>> do not match the subnets in that access list we set tag to 300. This route
>> map is then applied and the ISPs router sees only specific routes that have
>> a tag of 300 on them and other routes with no tags. The same thing is done
>> on Site B where the route map looks at an acl which contains networks only
>> at Site B and does nothing. But the 2nd statement sets tag to 200, meaning
>> all routes other than those specified in the acl are to be tagged as 200.
>>
>> What we want is for the ISP to match tag 300 and then make it more
>> expensive within their BGP cloud. The ISP redistributes EIGRP to BGP. Hence
>> the question, when redistributing from one protocol to another, are the tags
>> maintained or lost?
>>
>> The purpose of this exercies is that if remote sites want to access a
>> subnet at Site B they are correctly routed to the ISPs router at Site B and
>> do not go via their router at Site A and then via the dark fibre. This is
>> what we would like to achieve.
>>
>> Regards,
>>
>> Vikas.
>>
>> On Wed, Aug 19, 2009 at 10:25 AM, Adrian Brayton <[email protected]>wrote:
>>
>>> Thats a loaded question without enough detail...
>>>
>>> 1. Is the SP running MPLS where your EIGRP route's are added to a VRF?
>>>
>>> 2. Do you have multiple ISPs that you connect to?
>>>
>>> 3. Why do you want to make those routes more expensive in there cloud?
>>>
>>> 4. Are you running or going to be running BGP?
>>>
>>> Just a few more details would sure make answering this question a little
>>> easier!!
>>>
>>>
>>>
>>> On Aug 18, 2009, at 7:18 PM, Vikas Sharma wrote:
>>>
>>>   Hi Everyone,
>>>>
>>>> I need some help regarding redistribution of tagged routes. Issue is as
>>>> follows.
>>>>
>>>> I have a core switch on which is configured EIGRP. The ISP's router
>>>> connects into the core switch also runs EIGRP. The ISP in turn runs BGP
>>>> within their cloud.
>>>>
>>>> I have a distribute list on my core switch that sets a tag of 300 for
>>>> routes going towards the ISP router. That is fine and the ISP's router can
>>>> see them.
>>>>
>>>> The question I have is when they redistribute EIGRP into BGP are the
>>>> tags maintained? I want them to apply a route-map to match tag 300 and then
>>>> use the as-prepend command within the route-map to make those routes more
>>>> expensive within their cloud. The route-map then gets applied in the
>>>> outbound direction from within the router bgp (AS-Number).
>>>>
>>>> I would like to think there is no need to apply the route map in the
>>>> EIGRP config on the ISP router. Rather, first they redistribute EIGRP into
>>>> BGP and then apply this route-map in BGP. Please advice.
>>>>
>>>> Regards,
>>>>
>>>> --
>>>> Vikas Sharma
>>>> Network Specialist
>>>> Fujitsu Australia
>>>> (M): 0421 052 117
>>>> _______________________________________________
>>>> For more information regarding industry leading CCIE Lab training,
>>>> please visit www.ipexpert.com
>>>>
>>>
>>>
>>
>>
>> --
>> Vikas Sharma
>> Network Specialist
>> Fujitsu Australia
>> (M): 0421 052 117
>>
>> _______________________________________________
>> For more information regarding industry leading CCIE Lab training, please
>> visit www.ipexpert.com
>>
>>
>
>
> --
> Bryan Bartik
> CCIE #23707 (R&S), CCNP
> Sr. Support Engineer - IPexpert, Inc.
> URL: http://www.IPexpert.com
>



-- 
Vikas Sharma
Network Specialist
Fujitsu Australia
(M): 0421 052 117
_______________________________________________
For more information regarding industry leading CCIE Lab training, please visit 
www.ipexpert.com

Reply via email to