Just a side note - the problem of dynamic BGP peers is solvable,
depending on what you're running at the hub site.
http://www.cisco.com/en/US/docs/ios/iproute_bgp/configuration/guide/irg_neighbor.html#wp1055389
-- Marko Milivojevic - CCIE #18427 (SP R&S) Senior CCIE Instructor / Managing Partner - IPexpert On Fri, Feb 15, 2013 at 10:04 PM, Brian McGahan <bmcgahan_at_ine.com> wrote: > DMVPN scalability boils down to a hub-and-spoke routing logic problem with the other protocols besides EIGRP. Think legacy Frame-Relay hub-and-spoke, but at a much much larger scale. > > IS-IS can't be used for DMVPN because it lacks hub-and-spoke support. BGP could theoretically be used but it defeats the purpose of dynamic endpoints, which essentially is a show stopper for DMVPN. OSPF can be used, but in DMVPN Phase I or Phase II you'd have to run it as network type Broadcast because of the goal to preserve next-hop values between spoke advertisements. Phase III lifts this limitation and would allow you to run OSPF as point-to-multipoint, which helps with dual-hub or more than dual-hub designs, but it doesn't address the root of the problem. The real scalability limit with OSPF over DMVPN is the flooding scope. > > OSPF over DMVPN would require that all spokes participate in the same area - i.e. the flooding domain - as the hub(s). This implies that any flooding event initiated at one spoke site would impact all other spokes serviced by that hub. Incremental SPF could help somewhat here, but that doesn't fix the flooding problem, just the SPF run problem. Area hierarchy wouldn't really help you in this case, because OSPF only has two levels of hierarchy, area 0 and not area 0. At DMVPN scale numbers this means that you would have thousands or tens of thousands of routers in the same area. In reality it's not feasible. > > EIGRP on the other hand can form an arbitrary hierarchy based on both EIGRP Stub and prefix summarization. Both of these limit the EIGRP Query Domain, which is analogous to the OSPF flooding domain. This means that convergence events at one spoke site won't necessarily affect other stub sites. This scaling becomes even more apparent when the DMVPN network has more levels of hierarchy with regional hubs and central hubs, which again isn't really feasible with OSPF. > > > Brian McGahan, CCIE #8593 (R&S/SP/Security) > bmcgahan_at_INE.com > > Internetwork Expert, Inc. > http://www.INE.com > > -----Original Message----- > From: nobody_at_groupstudy.com [mailto:nobody_at_groupstudy.com] On Behalf Of Tony Singh > Sent: Friday, February 15, 2013 6:31 PM > To: Steve Di Bias > Cc: Cisco certification > Subject: Re: EIGRP offered as an RFC > > Suppose it's their science and they want to see what other vendors bring to the table, agree stub flags would need to be set but Cisco says "it will work with other vendors" so im sure it will be a a case of code implementation if a vendor implements this, also reading that it mentions that eigrp is the best solution for dmvpn scalability, I'm not that long in the tooth to agree or disagree here what else can be used for dmvpn solutions? > > Think in the end Cisco can only benefit and good luck > > -- > BR > > Tony > > Sent from my iPad > > On 16 Feb 2013, at 00:15, Steve Di Bias <sdibias_at_gmail.com> wrote: > >> I was quite shocked when I first heard this as well!! I'm not entirely >> sure > why the EIGRP stub feature is being left out though? Can someone please enlighten me? >> >> > http://www.cisco.com/en/US/prod/collateral/iosswrel/ps6537/ps6554/ps6599/ps66 > 30/qa_C67-726299.html >> >> Thanks, >> Steve Di Bias- CCIE #32840 >> >> >> On Fri, Feb 15, 2013 at 3:23 PM, Tony Singh <mothafungla_at_gmail.com> wrote: >> Hi >> >> Surprised to hear of this but guess it is better for Cisco in the end >> with a > long term SDN approach with Cisco One!? >> >> Notice they won't offer NSSA, thoughts... >> >> I heard a lecture from Brian Dennis mentioned he used it in certain SP > environments so still a use case it seems >> >> >> -- >> BR >> >> Tony >> >> Sent from my iPad >> >> >> Blogs and organic groups at http://www.ccie.net >> >> ______________________________________________________________________ >> _ Subscription information may be found at: >> http://www.groupstudy.com/list/CCIELab.html > > > Blogs and organic groups at http://www.ccie.net > > _______________________________________________________________________ > Subscription information may be found at: > http://www.groupstudy.com/list/CCIELab.html > > > Blogs and organic groups at http://www.ccie.net > > _______________________________________________________________________ > Subscription information may be found at: > http://www.groupstudy.com/list/CCIELab.html Blogs and organic groups at http://www.ccie.netReceived on Sat Feb 16 2013 - 10:47:30 ART
This archive was generated by hypermail 2.2.0 : Fri Mar 01 2013 - 07:57:58 ART