«

Hi,
I have a bunch of location connected to regional office. Regional office
summarize all routes to a central location. On central location there is
dialin router that handles location when they loose primary link and connect
to a that router on central location.
When location goes on backup, it sends its eigrp route son there is
summarized route from regional office and a specific one sent by location on
backup so traffic for that location is routed ok.
When primary of that location goes up, location starts to send packets thru
primary link. As isdn link is still up (because of delay timer) specific
eigrp route is still sent to a central location so during a transition time
we have asymetric routing - location sends packet trhu primary link, and
traffic is sent from central location thru isdn because of more specific
route.
Since traffic is not sent outside isdn interface on location (only received)
eventually location shuts down isdn interface.
But although isdn is down, eigrp neighborship between dialin router and
location via isdn is not immediately broken, so when traffic is sent from
central location to a remote location, specific route is still active in
dialin backup router. Immediately when packet enters that router, backup
router calls remote router and isdn connection is restored.
If there was no traffic during isdn going down and neighborship breaking,
route would dissapear and it would be fine. This way ISDN is very often up.
Solution: I could remove summarization done on regional office toward
central location so all specific routes would be visible on central
location. But design studies always recommend summarization in direction to
central location.

"jmiklo" <> wrote in message
news:cntlqj$84t$...
> Hi,
> I have a bunch of location connected to regional office. Regional office
> summarize all routes to a central location. On central location there is
> dialin router that handles location when they loose primary link and
connect
> to a that router on central location.
> When location goes on backup, it sends its eigrp route son there is
> summarized route from regional office and a specific one sent by location
on
> backup so traffic for that location is routed ok.
> When primary of that location goes up, location starts to send packets
thru
> primary link. As isdn link is still up (because of delay timer) specific
> eigrp route is still sent to a central location so during a transition
time
> we have asymetric routing - location sends packet trhu primary link, and
> traffic is sent from central location thru isdn because of more specific
> route.
> Since traffic is not sent outside isdn interface on location (only
received)
> eventually location shuts down isdn interface.
> But although isdn is down, eigrp neighborship between dialin router and
> location via isdn is not immediately broken, so when traffic is sent from
> central location to a remote location, specific route is still active in
> dialin backup router. Immediately when packet enters that router, backup
> router calls remote router and isdn connection is restored.
> If there was no traffic during isdn going down and neighborship breaking,
> route would dissapear and it would be fine. This way ISDN is very often
up.
> Solution: I could remove summarization done on regional office toward
> central location so all specific routes would be visible on central
> location. But design studies always recommend summarization in direction
to
> central location.
>
>

I've encountered a similar problem. I have designed backup connections
through broadband vpn. These connections are always up. Therefore, summary
routes don't work for me. You really can't summarize when you have
different entry points into the same network. That totally defeats the
whole purpose of summarization.

Here is a solution that may work for you... Start by removing the
summarization and just letting the whole route tables come to each location.
This will get you working right. Then, add an incoming distribute-list to
eigrp at the remote sites to just allow a default route. That way, you
don't have to worry about having large route tables at the remotes, but you
still will have a default route back to the main site. Your switchover
should be almost immediate when the sites come back up.

ex:
access-list 5 permit 0.0.0.0
access-list 5 deny any

router eigrp 100
distribute-list 5 in

Hope that helps,

Jim

> I've encountered a similar problem. I have designed backup connections
> through broadband vpn. These connections are always up. Therefore,
summary
> routes don't work for me. You really can't summarize when you have
> different entry points into the same network. That totally defeats the
> whole purpose of summarization.
>
> Here is a solution that may work for you... Start by removing the
> summarization and just letting the whole route tables come to each
location.
> This will get you working right. Then, add an incoming distribute-list to
> eigrp at the remote sites to just allow a default route. That way, you
> don't have to worry about having large route tables at the remotes, but
you
> still will have a default route back to the main site. Your switchover
> should be almost immediate when the sites come back up.
>
> ex:
> access-list 5 permit 0.0.0.0
> access-list 5 deny any
>
> router eigrp 100
> distribute-list 5 in
>
> Hope that helps,
>
> Jim
>
Thanks for sharing your experience. That way was also my idea (which I
didn't elaborate in details but goes with removing summarization toward
core). Central location (core) would have all specific location routes, but
would summarize back to regional offices and to end locations.
But I was wondering because summarization to the core is always recommended
in order to avoid route recalculation in core when some end locations goes
down.
Mostly all networks have centralized backup so all locations dials to the
same dialin routers so similar problem always exists.

thanks
j