From: Nick Shah (nshah@xxxxxxxxxxxxxx)
Date: Tue Jul 09 2002 - 23:14:03 GMT-3
> What is IPX NETWORK DISCOVERY? IS it a way to have an interface
> synamically gain an IPX network number?
Its a way to "figure out" the ipx network number of the other end, you have
to figure it out without doing a show on that other end, assuming you have
no administrative control over the other end. That other end could be a
server or a router.
> Also, If you use sub-interfaces instead of secondaries to support
> different encapsulations and your media is Ethernet, does this mean you
> need to create multiple VLANs for each sub-interace and segrate the A1
> network onto one "wire" and the B2 network onto another "wire"?
Not in this case. Basically the idea is that by putting all possible
encapsulations on subifs (or secondary networks) is that any one
encapsulation has to match the other end, and hence will "trigger" a
response (in terms of matching or not matching network). So assume that the
ethernet interface is directly connected to the Server/router/other end.
However, your Q does raise interesting observation. If the other end belongs
to a different VLAN, and you dont have isl/dot1q encapsulation configured on
ur end (including correct vlan), there is not an iota of chance to do
"network discovery".
You can also have 5 routers connected on a broadcast medium and have each
running different encaps, and one of the main router (whcih needs to speak
with eveyrone else) can run multiple encaps/multiple networks)
> If this were a lab I would say that would work but in real life the
> reason you might need one wore to have different networks and
> encapsulations because several were allready on one wire. Is this there
> a way to do the sub-interfaces where it is all still one big "wire"?
Once again, this is needed only to figure out the other end. Once you have
found the network on the other end, just configure it with the right
encapsulation (get rid of the others, they are put in only to discover the
network)
On another note, the behaviour is flaky, meaning I (including some other
people) couldnt get it to work at first try ( I just tried once). I will try
it on another set of routers running a different variant of ios.
rgds
Nick
>
>
>
>
> On 9 Jul 2002 15:15:43 -0000, "Prakash H Somani"
> <pdsccie@rediffmail.com> said:
> > Hi Nick....
> >
> > It works....find
> >
> > A. Debug output
> >
> > *Mar 6 04:09:29.498: IPXRIP: Marking network 222 FFFFFFFF for
> > Flash Update
> > *Mar 6 04:09:29.502: IPXRIP: General Query
> > src=222.00b0.64db.be20, dst=222.ffff
> > .ffff.ffff, packet sent (via Ethernet0/0)
> > *Mar 6 04:09:29.510: IPXRIP: positing flash update to
> > 22.ffff.ffff.ffff via Loo
> > pback0 (broadcast)
> > *Mar 6 04:09:29.510: IPXRIP: positing flash update to
> > 222.ffff.ffff.ffff via Et
> > hernet0/0 (broadcast)
> > *Mar 6 04:09:29.514: IPXRIP: positing full update to
> > 222.ffff.ffff.ffff via Eth
> > ernet0/0 (broadcast)
> > *Mar 6 04:09:29.566: IPX:
> > Lo0:22.0002.0002.0002->22.ffff.ffff.ffff ln= 40 tc=00
> > pt=01 ds=0453 ss=0453, rcvd
> > *Mar 6 04:09:29.566: IPX:
> > Lo0:22.0002.0002.0002->22.ffff.ffff.ffff ln= 40 tc=00
> > pt=01 ds=0453 ss=0453, bad pkt
> > *Mar 6 04:09:29.566: IPXRIP: suppressing null update to
> > 222.ffff.ffff.ffff (Eth
> > ernet0/0)
> > R2(config-if)#
> > *Mar 6 04:10:14.928: IPX:
> > Et0/0:5.00d0.58ad.27f1->5.ffff.ffff.ffff ln= 40 tc=00
> > pt=01 ds=0453 ss=0453, rcvd
> > *Mar 6 04:10:14.928: IPX:
> > Et0/0:5.00d0.58ad.27f1->5.ffff.ffff.ffff ln= 40 tc=00
> > pt=01 ds=0453 ss=0453, bad pkt
> > *Mar 6 04:10:27.429: IPXRIP: positing full update to
> > 22.ffff.ffff.ffff via Loop
> > back0 (broadcast)
> > *Mar 6 04:10:27.437: IPX:
> > Lo0:22.0002.0002.0002->22.ffff.ffff.ffff ln= 40 tc=00
> > pt=01 ds=0453 ss=0453, rcvd
> > *Mar 6 04:10:27.437: IPX:
> > Lo0:22.0002.0002.0002->22.ffff.ffff.ffff ln= 40 tc=00
> > pt=01 ds=0453 ss=0453, bad pkt
> > *Mar 6 04:10:29.348: IPXRIP: positing full update to
> > 222.ffff.ffff.ffff via Eth
> > ernet0/0 (broadcast)
> >
> > B. Configuration details
> >
> > R5#sh runn int e0
> > Building configuration...
> >
> > Current configuration:
> > !
> > interface Ethernet0
> > ip address 150.50.50.2 255.255.255.240 secondary
> > ip address 172.16.45.5 255.255.255.248
> > no ip directed-broadcast
> > no ip split-horizon eigrp 100
> > ipx network 5 encapsulation SAP
> > end
> >
> > R2#sh run int e0/0
> > Building configuration...
> >
> > Current configuration:
> > !
> > interface Ethernet0/0
> > ip address 172.16.45.2 255.255.255.248
> > no ip directed-broadcast
> > ipx network 222 encapsulation SAP
> > end
> >
> > regards....Prakash
> >
> > On Tue, 09 Jul 2002 Nick Shah wrote :
> > >There was a post in the archives, which allowed for a method
> > >(coupled with
> > >debug ip packet) to find the IP EIGRP AS ##, but nothing that I
> > >know of for
> > >IPX eigrp.
> > >
> > >Btw, I juggled (read struggled :) a bit trying to get the ipx
> > >network
> > >discovery.. I tried the following but somehow couldnt get it to
> > >work...
> > >
> > >* create sub-ifs with different encapsulations, and assign
> > >different ipx
> > >networks to it
> > >* turn on debug ipx sap , debug ipx packet, debug ipx routing (I
> > >think)
> > >*but it showed me everything except what I wanted (bad IPX
> > >network etc.
> > >etc.)
> > >
> > >I used secondary ipx addresses (in lieu of sub interfaces), used
> > >ipx eigrp,
> > >ipx rip & even NLSP.. but no luck..
> > >
> > >rgds
> > >Nick
> > >----- Original Message -----
> > > From: "Todd Carswell" <acarswell@nc.rr.com>
> > >To: <ccielab@groupstudy.com>
> > >Sent: Tuesday, July 09, 2002 1:23 PM
> > >Subject: IPX EIGRP Autonomous System Discovery -- Is it
> > >possible?
> > >
> > >
> > > > Hi all,
> > > >
> > > > I can do IPX NETWORK discovery. It made me wonder...
> > > >
> > > > Is it possible to discover a neighboring IPX EIGRP Autonomous
> > >System
> > >Number?
> > > >
> > > > Thx
> > > >
> > > > Todd Carswell
> > > >
This archive was generated by hypermail 2.1.4 : Sat Sep 07 2002 - 19:36:24 GMT-3