Rule 2 is always true for IPv4 for all the protocols (IPV6 does not do it by default). Even if you are redistributing OSPF to BGP, you will get the connected routes redistributed. Remember that "redistribute connected" is an independent command in itself and has nothing to do with redistributing other protocols.
Thanks On Thu, Jun 12, 2008 at 9:48 PM, Carlos Valero <[EMAIL PROTECTED]> wrote: >> Redistribution from any protocol will include all the connected interfaces >> on which that protocol is active. > > Thank you; I did not know that "rule". I guess that explains it. > > Regarding the second part of this (redistribute connected), that doesn't > apply when redistributing OSPF routes into BGP, since the only match options > are: > > - internal > - external & > - nssa > > > CV. > > > --- On Thu, 6/12/08, Suresh Mishra <[EMAIL PROTECTED]> wrote: > > From: Suresh Mishra <[EMAIL PROTECTED]> > Subject: Re: [OSL | CCIE_RS] Redistribution of OSPF Routes into BGP > To: "OSL CCIE Routing and Switching Lab Exam" <[email protected]> > Date: Thursday, June 12, 2008, 2:41 PM > > Remember the rules below for redistribution. > > > For IPV4 > > 1. Redistribution from any protocol will inlcude all the connected > interfaces on which that protocol is active. In addition to that all > the routes that will show up using the " show ip route " command on > that router will also be redistributed. > 2. However, if you use "redistribute connected" then the rule will > change. The redistribute connected will govern the > redistribution of connected interfaces. > > For OSPF :- Internal routes are redistributed by default. > > For BGP :- Internal routes (ibgp) are not redistributed by default. > Only ebgp routes are redistributed. > > Suresh > > > > > On Thu, Jun 12, > 2008 at 10:11 AM, Roger RPF <[EMAIL PROTECTED]> wrote: >> Hi Carlos, >> >> >> >> I just made a short check on my stack. I have the following, partial > setup: >> >> >> >> ----<OSPF R6 EIGRP>------<EIGRP R9 >--- >> >> >> >> The interface on the OSPF side on R6 is configured in OSPF with a network >> statement. Now, on R6 I do a redistribute ospf 1 match internal under my >> EIGRP process. If I check the routing table on R9, I can see that the > prefix >> got redistributed: >> >> >> >> D EX 170.1.100.0/30 [170/40537600] via 170.1.69.1, 00:05:40, > Serial0/2/0 >> >> >> >> So, to me, this means that it takes the connected when you specify match >> internal. >> >> >> >> regards >> >> >> >> Roger >> >> >> >> Von: [EMAIL PROTECTED] >> > [mailto:[EMAIL PROTECTED] Im Auftrag von Carlos Valero >> Gesendet: Donnerstag, 12. Juni 2008 15:12 >> An: OSL CCIE Routing and Switching Lab Exam >> Betreff: Re: [OSL | CCIE_RS] Redistribution of OSPF Routes into BGP >> >> >> >> Hello, >> >> I would like to confirm/clarify an issue about OSPF redistribution that > may >> seem basic, but it's still confusing me a little bit. If we have the >> following statements: >> >> router bgp 1 >> redistribute ospf 1 match internal >> >> Question is: Does "match internal" imply internal OSPF routes + > "connected >> routes" ? >> >> I've read a few articles about it, and although none confirms (or > denies) >> that, the examples I've seen, seem to confirm that connected routes > are also >> redistributed with the "internal" keyword (or actually by > default). >> >> For > instance, the following link provides an exercise in which connected >> routes are in fact redistributed with the "internal" keyword: >> > http://www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a00800943c5.shtml >> >> And this seems to happen in our Labs as well. For instance, in Lab 7 > (v9.0) >> Task 6, we have the following scenario: >> >> R7 & R8 are connected through a Serial Link and BGP + OSPF are > configured >> over that link: >> >> R8 >> >> router bgp 65078 >> neighbor 150.50.5.69 remote-as 65078 >> >> router ospf 1 >> network 150.50.5.69 0.0.0.0 area 0 >> >> R7 >> >> router bgp 65078 >> neighbor 150.50.5.68 remote-as 65078 >> >> router ospf 1 >> network 150.50.5.68 0.0.0.0 area 0 >> >> If I do a "sh ip route" on either router, of course > 150.50.5.64/27 appears >> as a "connected" > route; not an OSPF route. >> >> If I do a "sh ip bgp" on R7, 150.50.5.64/27 does NOT show up > there. >> >> So up to this point, 150.50.5.64/27 is not an OSPF route, nor a BGP > route. >> But if I do OSPF redistribution: >> >> router bgp 65078 >> redistribute ospf 1 match internal >> >> If I now do a "sh ip bgp" on R7, 150.50.5.64/27 now appears > there! >> >> Actually, part of this task requires us to filter this route out, so we > need >> to apply a route-map to make sure this route doesn't get > redistributed! >> >> So again, the question is: >> >> Does "match internal" imply internal OSPF routes + > "connected routes" ? >> >> If that's not the case, then why 150.50.5.64/27 gets redistributed > with >> "redistribute ospf 1 match internal" >> >> Thank you! >> >> >> C. Valero. >> --- >> >> > __________________________________________________ >> Do You Yahoo!? >> Tired of spam? Yahoo! Mail has the best spam protection around >> http://mail.yahoo.com >
