|
Thread Index Re:[isp-bgp] isp-bgp digest: November 16, 2005
Hi I haven't read that url yet, but "Normal" setups would have BGP peer directly connected, but with with the multihop (loopback ip) resulting in physical interface of immediate peer becoming the intermediate hop. I can't imagine designing a bgp network where peers are remote. Maybe gre tunnel for some quirky minds... Any ideas (scenarios) where bgp peer is remote and an intermediate router is the bone in the kebab? Rgrds On 11/17/05, ISP-BGP Discussion List digest <isp-bgp@...> wrote: > ISP-BGP Digest for Wednesday, November 16, 2005. > > 1. Re:BGP multihop > 2. Re: BGP multihop > 3. Re: BGP multihop > 4. Re: BGP multihop > 5. Re: BGP multihop > > ---------------------------------------------------------------------- > > Subject: Re:BGP multihop > From: vdadlaney@... > Date: Wed, 16 Nov 2005 14:22:27 +0000 > X-Message-Number: 1 > > Hi All, > > Just to add I believe that you need a specific route to the ip you specify in the neighbor statement. Usually EBGP is configured via a directly connected link to the neighbor but in cases where the neighbor might not be directly connected you would have to specify ebgp multiphop since the ttl on EBGP is 1 by default. You cannot use a default route for the peering relationship. It has to be specific. Please correct me if I am wrong. Thanks > > Regards, > > Vikram > > > > > Michal.Grzybczyk@vattenfall.pl wrote: > > > > I have problem with multihop configured with eBGP. > > > > Is it normal that I need configured routes on 'intermediate' > > > > router to each destination networks I've received through BGP ? > > > > To each desitnation network?? NO. To recap what Pete said, the > > static/IGP routes are required pointing to eBGP routers, so that they > > can establish peering sessions and exchange routes... > > > > > > > > BGP requires that its neighbors be reachable through non-BGP routes. > > > Otherwise, if/when the routes disappear, it can never determine > > > reachability to the neighbor without reachability to the neighbor. > > > > > > The alternative is static or dynamic IGP routes to the neighbor. > > > > > > pt > > > > > > > -- > > Nonchalantly yours > > GobbledeGeek > > [Everything but Gobbledegook.. !!] > > > > To unsubscribe via postal mail, please contact us at: > > Jupitermedia Corp. > > Attn: Discussion List Management > > 475 Park Avenue South > > New York, NY 10016 > > > > Please include the email address which you have been contacted with. > > > > Copyright 2005 Jupitermedia Corporation All Rights Reserved. > > ---------------------------------------------------------------------- > > Subject: Re: BGP multihop > From: Ton S <ton-s-lijst@...> > Date: Wed, 16 Nov 2005 09:40:48 -0500 > X-Message-Number: 2 > > ANY route will do. Even Default Gateway, or any other dynamic routing > protocol. > Heck, even a BGP route (from a different BGP process, but Cisco will not > let you code more than one BGP instance per router) could do the trick. > > as long as the route covers the destination, is valid, has a reverse > path, the BGP session would come up > > Ton > > vdadlaney@... wrote: > > >Hi All, > > > >Just to add I believe that you need a specific route to the ip you specify in the neighbor statement. Usually EBGP is configured via a directly connected link to the neighbor but in cases where the neighbor might not be directly connected you would have to specify ebgp multiphop since the ttl on EBGP is 1 by default. You cannot use a default route for the peering relationship. It has to be specific. Please correct me if I am wrong. Thanks > > > >Regards, > > > >Vikram > > > > > > > > > ---------------------------------------------------------------------- > > Subject: Re: BGP multihop > From: vdadlaney@... > Date: Wed, 16 Nov 2005 15:53:21 +0000 > X-Message-Number: 3 > > Hi Ton, > > This is a quote directly from cisco's website. > > " To avoid the accidental creation of loops through oscillating routes, the multihop session will not be established if the only route to the multihop peer's address is the default route (0.0.0.0). " > > Here's the link. Watch the word wrap or you can google "configuring bgp" and than search for the text above in the document. > > http://www.cisco.com/univercd/cc/td/doc/product/software/ios120/12cgcr/np1_c/1cprt1/1cbgp.htm > > > Thanks > > Regards, > > Vikram > > > > ANY route will do. Even Default Gateway, or any other dynamic routing > > protocol. > > Heck, even a BGP route (from a different BGP process, but Cisco will not > > let you code more than one BGP instance per router) could do the trick. > > > > as long as the route covers the destination, is valid, has a reverse > > path, the BGP session would come up > > > > Ton > > > > vdadlaney@... wrote: > > > > >Hi All, > > > > > >Just to add I believe that you need a specific route to the ip you specify in > > the neighbor statement. Usually EBGP is configured via a directly connected link > > to the neighbor but in cases where the neighbor might not be directly connected > > you would have to specify ebgp multiphop since the ttl on EBGP is 1 by default. > > You cannot use a default route for the peering relationship. It has to be > > specific. Please correct me if I am wrong. Thanks > > > > > >Regards, > > > > > >Vikram > > > > > > > > > > > > > > > To unsubscribe via postal mail, please contact us at: > > Jupitermedia Corp. > > Attn: Discussion List Management > > 475 Park Avenue South > > New York, NY 10016 > > > > Please include the email address which you have been contacted with. > > > > Copyright 2005 Jupitermedia Corporation All Rights Reserved. > > ---------------------------------------------------------------------- > > Subject: Re: BGP multihop > From: Ton S <ton-s-lijst@...> > Date: Wed, 16 Nov 2005 11:08:42 -0500 > X-Message-Number: 4 > > Ahh, and how does the peering router determine wether any intermediate > non-BGP router (and that's what we were talking about, not directly > connected BGP peers) has a default route? > > Ton > > vdadlaney@... wrote: > > >Hi Ton, > > > >This is a quote directly from cisco's website. > > > >" To avoid the accidental creation of loops through oscillating routes, the multihop session will not be established if the only route to the multihop peer's address is the default route (0.0.0.0). " > > > >Here's the link. Watch the word wrap or you can google "configuring bgp" and than search for the text above in the document. > > > >http://www.cisco.com/univercd/cc/td/doc/product/software/ios120/12cgcr/np1_c/1cprt1/1cbgp.htm > > > > > >Thanks > > > >Regards, > > > >Vikram > > > > > > > ---------------------------------------------------------------------- > > Subject: Re: BGP multihop > From: vdadlaney@... > Date: Wed, 16 Nov 2005 16:18:15 +0000 > X-Message-Number: 5 > > Hi Ton, > > Ah missed that keyword "intermediate" in the original request. Thx for pointing tha t out. Apologize for the error folks. Thx > > Regards, > > Vikram > > > > Ahh, and how does the peering router determine wether any intermediate > > non-BGP router (and that's what we were talking about, not directly > > connected BGP peers) has a default route? > > > > Ton > > > > vdadlaney@... wrote: > > > > >Hi Ton, > > > > > >This is a quote directly from cisco's website. > > > > > >" To avoid the accidental creation of loops through oscillating routes, the > > multihop session will not be established if the only route to the multihop > > peer's address is the default route (0.0.0.0). " > > > > > >Here's the link. Watch the word wrap or you can google "configuring bgp" and > > than search for the text above in the document. > > > > > >http://www.cisco.com/univercd/cc/td/doc/product/software/ios120/12cgcr/np1_c/1c > > prt1/1cbgp.htm > > > > > > > > >Thanks > > > > > >Regards, > > > > > >Vikram > > > > > > > > > > > > To unsubscribe via postal mail, please contact us at: > > Jupitermedia Corp. > > Attn: Discussion List Management > > 475 Park Avenue South > > New York, NY 10016 > > > > Please include the email address which you have been contacted with. > > > > Copyright 2005 Jupitermedia Corporation All Rights Reserved. > > > > --- > > END OF DIGEST > > --- > You are currently subscribed to isp-bgp as: gobbledegeek@... > To unsubscribe send a blank email to leave-isp-bgp-1708663D@... > To unsubscribe via postal mail, please contact us at: > Jupitermedia Corp. > Attn: Discussion List Management > 475 Park Avenue South > New York, NY 10016 > > Please include the email address which you have been contacted with. > > Copyright 2005 Jupitermedia Corporation All Rights Reserved. > -- Nonchalantly yours GobbledeGeek [Everything but Gobbledegook.. !!] To unsubscribe via postal mail, please contact us at: Jupitermedia Corp. Attn: Discussion List Management 475 Park Avenue South New York, NY 10016 Please include the email address which you have been contacted with. Copyright 2005 Jupitermedia Corporation All Rights Reserved. <- Previous Message | Next Message -> Thread Index |
|
