From: Marc La Porte (marc.a.laporte@gmail.com)
Date: Thu Feb 28 2008 - 14:54:40 ARST
Agree. It's more a precaution. If however though you would (in a future
task) need to configure an IP address on the main interface you are not
running into any surprises... Unless otherwise mentioned, I would always do
an "no frame inv" just to avoid surprise (like those nasty mappings to
0.0.0.0)
On Thu, Feb 28, 2008 at 5:19 PM, Andy <and123and@googlemail.com> wrote:
> aha, but, as there is no IP address on either main serial interface of R1
> or R6, then neither is going to :
> a) send inverse-arp as there is no IP address to inverse-arp for, and
> b) reply to inverse-arp as there is no one sending a request on this DLCI
> pair
>
> So, by putting the "no fr inv arp" command on we are not actually
> acheiving anything that was not there before, the question said "ensure that
> DLCIs 110 and 104 between R1 and R6 are not used", I can not see what is
> gained by adding the command. :-/
>
>
>
>
>
> On 28/02/2008, Marc La Porte <marc.a.laporte@gmail.com> wrote:
> >
> > Yes, so as there are multiple DLCIs available on the main interface, and
> > only one of those is used on the p2p sub-interface (which by default doesn't
> > do I-ARP), it would make sense to use the "no frame inv" on the main
> > interface to ensure no dynamic mapping between R1 and R6 is created via
> > Inverse-ARP. You could also do this by moving all the non-used DLCIs to an
> > unused multipoint sub-interface...
> >
> > Indeed, the Virtual-Interface as such has no influence on the InverseARP
> > process.
> >
> > HTH.
> > Marc
> >
> > On Thu, Feb 28, 2008 at 4:43 PM, Andy <and123and@googlemail.com> wrote:
> >
> > > For those that have the book, the question and diag are on p77. Lab 2.
> > >
> > > For those that dont: Hub and spoke with R4 as the hub.
> > >
> > > R1 has DLCI 101 and 110
> > > R4 has DLCI 100 and 102
> > > R6 has DLCI 104 and 103
> > >
> > > Mappings are
> > > 101 - 100
> > > 110 - 104
> > > 103 - 102
> > >
> > > I hope you can follow this, the question is:
> > >
> > > - Config the FR portion of the network as shown in figure x and ensure
> > > that DLCIs 110 and 104 between R1 and R6 are not used.
> > > - Use p2p subinterfaces
> > >
> > > The solution shows IP address on p2p and "no fram inverse-arp" on the
> > > main interface.
> > >
> > > The eventual solution uses PPP on the p2p interfaces with a virtual
> > > interface but I mention it here only as reference, I do not think it matters
> > > as to why we are using "no fr inv-arp" on main interface.
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > R1 (ser0/1 DLCI 101) to R4 (ser0/0 DLCI 100)
> > > R1 (ser0/1 DLCI 110) to R6 (ser5/0 DLCI 104)
> > > R6 (ser5/0 DLCI 103) to R4 (ser0/0 DLCI
> > >
> > >
> > > On 28/02/2008, Marc La Porte <marc.a.laporte@gmail.com> wrote:
> > > >
> > > > Andy,
> > > >
> > > > Are both DLCIs 104 and 110 on both R1 and R6? And the p2p
> > > > sub-interface is for instance using DLCI 104 on both sides? Then the "no
> > > > frame inv" makes sense on the main interface, because all DLCIs coming in
> > > > from the FR switch on that main interface are automatically assigned to the
> > > > main interface, so if you're using DLCI 104 (for instance) on your p2p
> > > > sub-interface that still leaves you with DLCI 110 "assigned" to the main
> > > > interface. Granted, as long as there is no IP address on the main interface
> > > > there is no I-ARP going on (I listen to the Brian's too --- or should I say
> > > > "the Brains"), but if you want to be sure then that's the way to go...
> > > >
> > > > If you could gives us a copy of the config that would help to (like
> > > > you said) something subtle...
> > > >
> > > > Marc
> > > >
> > > > On Thu, Feb 28, 2008 at 3:43 PM, Andy <and123and@googlemail.com>
> > > > wrote:
> > > >
> > > > > Hi
> > > > >
> > > > > On pages 92 and 93 of CCIE Routing and Switching LAB Practice labs
> > > > > by Cisco
> > > > > it gives a solution which uses "no frame relay inverse-arp" after
> > > > > asking the
> > > > > question "ensure that DLCIs 110 and 104 between R1 and R6 are not
> > > > > used. The "no fr inverse-arp" is configed on a main interface and
> > > > > a
> > > > > sub-interface used for p2p connection (with the IP going on the
> > > > > sub-interface).
> > > > >
> > > > > Now I know that "no frame relay inverse-arp" on an interface
> > > > > without an IP
> > > > > address does nothing (coz I have read the mails from Brian Dennis
> > > > > ;-) But
> > > > > then why does the Cisco Press book award one point for this??? In
> > > > > a lab
> > > > > scenario I would not config the command where an IP address does
> > > > > not exist
> > > > > as I would view this as overconfiguration. Am I missing something
> > > > > subtle
> > > > > here?
> > > > >
> > > > > -A
> > > > >
> > > > >
> > > > > _______________________________________________________________________
> > > > > Subscription information may be found at:
> > > > > http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Sat Mar 01 2008 - 16:54:50 ARST