Re: OSPF Problem

Just to add to some of the points that others have stated so nicely:

I think a GRE is closer to sham-link than a virtual link. But this can be
argued either way.

When your CEs are configured in area 0 separated by the MPLS cloud, you have
few choices:

   - A GRE configured on the PEs can do the trick
   - A sham link can also do the trick
   - If the cloud needs to be used as the primary path, and neither GRE nor
   sham-link is configured, dont configure the backup path in area 0,
   configure it in another area and then manipulate the cost, you may need a
   virtual-link to connect two aea 0s to each other through the backup link,
   incase the cloud is down.
   - If the backup link is also configured in area 0, then a sham-link/GRE
   through the cloud with some manipulation of cost on the backup link will
do
   the trick.
   - Virtual link is used for connection of non-zero area/s to area 0, or
   when we have a fragmented area 0, or when we want to avoid an area 0
   fragmentation through another area.
   - Sham link can extend any area.

On Tue, Jul 5, 2011 at 9:50 AM, Brian McGahan <bmcgahan_at_ine.com> wrote:

> It is an extension of the area. It extends the flooding domain,
> which in turn changes the route calculation process. Not only does this
> modify which paths will be preferred, but it also affects how changes in
the
> OSPF topology at one VPN site will affect the others.
>
> Normally the MPLS network is considered the "super backbone" for
> OSPF, because even if the PE-CE links are running OSPF area 0, the flooding
> domain is not extended over the MPLS network. This means from the
> perspective of the CE or C routers, the PEs look like ABRs. They originate
> either Type 3 Network Summary LSAs, Type 4 ASBR Summary LSAs, or Type 5
> External LSAs to describe other VPN sites.
>
> The key is that in OSPF design, the ABR is in charge of topology
> summarization. If a change in the OSPF topology occurs at VPN site 1, it
> will not necessarily result in an SPF run at other VPN sites. Instead it
> results in what's known as a Partial SPF or Partial Route Calculation
(PRC).
> This is why OSPF is sometimes described as distance vector between areas,
> while it is only link-state inside the area. When a sham-link is used to
> extend the area over MPLS though, this behavior changes.
>
> Per the RFC "The sham link is an unnumbered point-to-point
> intra-area link and is advertised as a type 1 link in a type 1 LSA." This
> means that the sham-link is considered a normal portion of the SPF graph of
> the area, and must be used for flooding. Like a virtual-link, the
sham-link
> runs as a demand circuit. This means that periodic flooding will not
occur,
> but normal flooding must occur if there is a change in the topology.
>
> The end result of this is that sham-links, just like virtual-links,
> limit the scalability of the network. If you were to run sham-links
between
> all of your sites it would basically be the same as running one flat area
> everywhere. Changes anywhere in the topology will require an SPF run, or
at
> least an iSPF run, everywhere in the network.
>
>
>
> Brian McGahan, CCIE #8593 (R&S/SP/Security)
> bmcgahan_at_INE.com
>
> Internetwork Expert, Inc.
> http://www.INE.com <http://www.ine.com/>
>
>
> -----Original Message-----
> From: nobody_at_groupstudy.com [mailto:nobody_at_groupstudy.com] On Behalf Of
> -Hammer-
> Sent: Tuesday, July 05, 2011 11:00 AM
> To: Routing Freak
> Cc: ccielab_at_groupstudy.com
> Subject: Re: OSPF Problem
>
> It's probably not the best answer but I hope it helps. Lastly, you keep
> referencing (for the sham link) extending a non-backbone to the backbone.
> I'm pretty sure a sham link is between two routers regardless of backbone
> state. In other words, it's not like a virtual link where your whole
purpose
> is to get to area 0. In a sham link, you are just making routes available
> between two PE routers regardless of where area
> 0 is. Sham links are not depending on particular OSPF areas as much as they
> are tied to specific OSPF processes.
>
> Anyone jump in if I'm misunderstanding this.
>
> -Hammer-
>
> "I was a normal American nerd"
> -Jack Herer
>
>
>
> On 07/05/2011 10:56 AM, Routing Freak wrote:
> > hmm Thanks Hammer for ur valuable info
> >
> > I m looking for some GREAT's response
> >
> > On Tue, Jul 5, 2011 at 9:23 PM, -Hammer- <bhmccie_at_gmail.com
> > <mailto:bhmccie_at_gmail.com>> wrote:
> >
> > There are certainly more qualified folks here that may speak up
> > but I do not look at it as extending a non backbone link. I look
> > at it as a method to enforce more preferred (the cloud) routing
> > between PEs. I look at it more or less as a pvc or a tunnel
> > between the PEs. I'm sure that's not the best way but it works for
> > me to think thru it.
> >
> > -Hammer-
> >
> > "I was a normal American nerd"
> > -Jack Herer
> >
> >
> >
> >
> > On 07/05/2011 10:48 AM, Routing Freak wrote:
> >> Hey Hammer, Good one
> >>
> >> But stilll I hava a problem, Is sham link extending ur non
> >> backbone link to the backbone link ( ie ur Super Backbone)
> >>
> >>
> >> On Tue, Jul 5, 2011 at 9:12 PM, -Hammer- <bhmccie_at_gmail.com
> >> <mailto:bhmccie_at_gmail.com>> wrote:
> >>
> >> Dude I didn't mean to simplify it. You were asking for the
> >> difference and I thought a clearer explanation might have
> >> helped. I don't look at these as being in the same realm and
> >> that's why maybe I am missing where you are confused. Yes,
> >> V-Links are an "extension" of area 0. Whereas sham links are
> >> used to "bridge" together multiple PEs in the same MPLS VPN
> >> backbone. The sham link gets you past the default OSPF rules
> >> for selecting intra-area routes instead of inter-area. The
> >> sham link also allows the sites to communicate over the MPLS
> >> VPN instead of via any external paths not provided via the
> >> cloud. Does that help?
> >>
> >> -Hammer-
> >>
> >> "I was a normal American nerd"
> >> -Jack Herer
> >>
> >>
> >>
> >>
> >> On 07/05/2011 10:33 AM, Routing Freak wrote:
> >>> Dude i know how to configure ospf sham link . I need to know
> >>> about the difference technically speaking differences
> >>>
> >>> Share some theory behind both, What r the similarities and
> >>> what r the differences?
> >>>
> >>>
> >>> On Tue, Jul 5, 2011 at 8:55 PM, -Hammer- <bhmccie_at_gmail.com
> >>> <mailto:bhmccie_at_gmail.com>> wrote:
> >>>
> >>> This helped me understand sham links better a while
> back....
> >>>
> >>>
> >>> http://blog.ipexpert.com/2010/01/20/introduction-to-ospf-sham-link/
> >>>
> >>> -Hammer-
> >>>
> >>> "I was a normal American nerd"
> >>> -Jack Herer
> >>>
> >>>
> >>>
> >>> On 07/05/2011 10:17 AM, Routing Freak wrote:
> >>> > Hey mates,
> >>> > I have a little confusion. What is the difference
> >>> between Sham link and
> >>> > Virtual Link?
> >>> >
> >>> > I know that Virtual Link is to extend ur backbone
> >>> > Sham link is used as a false link between two PE's .
> >>> Is it extending ur
> >>> > Area to other area and fool them that they r in the
> >>> same area??
> >>> >
> >>> > Because Sham link mainly used as a virtual cable which
> >>> sends Intra area
> >>> > LSA's ( which is the Type 1 LSA's) to other PE .
> >>> >
> >>> > What does exactly the difference betwen Virtual links
> >>> and Sham links
> >>> >
> >>> >
> >>> > 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.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
>
>
>
>
>
>
>
>

--
*Narbik Kocharians
*CCSI#30832, CCIE# 12410 (R&S, SP, Security)
www.MicronicsTraining.com <http://www.micronicstraining.com/>
Sr. Technical Instructor
*Ask about our FREE Lab Voucher with our Boot Camps*
YES! We take Cisco Learning Credits!
Training & Remote Racks available
Blogs and organic groups at http://www.ccie.net