From: yu chunyan (yuchunyan@hotmail.com)
Date: Mon Jul 28 2003 - 21:29:02 GMT-3
unfortunately. This is true just if one of two protocols is Eigrp. bc eigrp
have different distance for internal and external. you can study your first
diagram carefully. you will find that is the point to stop the route loop
without filter.
Bin Lei.
>From: "ccie2be" <ccie2be@nyc.rr.com>
>Reply-To: "ccie2be" <ccie2be@nyc.rr.com>
>To: "Group Study" <ccielab@groupstudy.com>
>Subject: Routing Protocols which tag External routes & route filters
>Date: Mon, 28 Jul 2003 08:38:47 -0400
>
>Hi,
>
>I'd like to verify something. It seems to me that when doing redist
>between
>routing protocols which tagged redistributed routes as External, like OSPF
>and
>Eigrp, it's not necessary to configure route filters to prevent routes
>which
>have been redist from being redist back into the their original routing
>protocol.
>
>For example, if the following topology exists
>
>
> EIGRP domain
>
> ---------R1-----------------R2-------------
>
> OSPF domain
>
>
>and 2-way redistribution is configured at both R1 and R2, there's no need
>for
>route filters because an OSPF route which is redist by R1 into the Eigrp
>domain is marked as External and therefore will not be redistributed back
>into the OSPF domain at router R2. In other words, routers doing redist,
>by
>default, filter routes tagged as External. What I'd like to confirm is
>that
>this is true in all cases, not just the ones I've come across ( I've done 2
>or
>3 practice labs where this was true).
>
>Also, I'd like to confirm that if the above is true, then the following is
>also true.
>
>
> RIP domain
>
>---------R1-----------------R2-------------
>
> Eigrp domain
>
>
>
>In this situation, route filters are only needed when redist RIP into Eigrp
>on
>R1 and R2 because when an Eigrp is redist into RIP, RIP doesn't distinquish
>between "native" RIP routes and routes redist into RIP and therefore would
>redist those routes back into the Eigrp domain unless a route filter were
>configured to prevent that.
>
>I'm hoping someone in Group Study will either confirm this theory or debunk
>it
>and explain why this is true or false.
>
>Thanks in advance, Raj
>
>
>_______________________________________________________________________
>You are subscribed to the GroupStudy.com CCIE R&S Discussion Group.
>
>Subscription information may be found at:
>http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Wed Aug 06 2003 - 06:52:56 GMT-3