All,I don't have the exact config, but here is an equivalent of the config that 
did not work.  However, there was full ip connectivity between all routers, 
including self-ping.  As I said, the symptom was seen on R3 where it never saw 
a "key 2" from R1, only a key 1.  In the real lab, I added send and accept 
lifetimes on all keys, so all were valid.  Nieghbours came up between R1 and R2 
fine, R1 thought that R3 was a nieghbour, but not vice versa.  
R1 (hub)
 
key chain EIGRP
key 1
key-string TOR2
key 2
key-string TOR3
 
int ser0/1/0
encap frame
no frame inverse
ip address 10.0.0.1 255.255.255.0
frame map ip 10.0.0.1 102 broad
frame map ip 10.0.0.2 102
frame map ip 10.0.0.3 103 broad
ip authentication mode eigrp 1 md5
ip authentication key-chain eigrp 1 EIGRP
 
router eigrp 1
no auto
net 10.0.0.0
 
+++++
 
key chain EIGRP
key 1
key-string TOR2
 
R2 (spoke 1)
 
int ser0/1/0
encap frame
no frame inverse
ip address 10.0.0.2 255.255.255.0
frame map ip 10.0.0.1 201 broad
frame map ip 10.0.0.2 201 
frame map ip 10.0.0.3 201
ip authentication mode eigrp 1 md5
ip authentication key-chain eigrp 1 EIGRP
 
router eigrp 1
no auto
net 10.0.0.0
 
+++++++++++++++++++++++++++
 
R3 (spoke 2)
 
key chain EIGRP
key 2
key-string TOR3
 
int ser0/1/0
encap frame
no frame inverse
ip address 10.0.0.3 255.255.255.0
frame map ip 10.0.0.1 301 broad
frame map ip 10.0.0.2 301 
frame map ip 10.0.0.3 301
ip authentication mode eigrp 1 md5
ip authentication key-chain eigrp 1 EIGRP
 
router eigrp 1
no auto
net 10.0.0.0

 > CC: [email protected]; [email protected]
> From: [email protected]
> Subject: Re: [OSL | CCIE_RS] EIGRP Authentication on NBMA Hub and Spoke
> Date: Thu, 3 May 2012 09:08:23 -0500
> To: [email protected]
> 
> I agree with one key at a time.
> 
> I imagine his GRE was Lo to Lo source/destination since he didn't mention IP 
> Unnum.
> 
> George-
> Can you show an example of your config? I speak IOS better than word 
> problems. Hopefully I speak for a large number of Engineers on here. ;)
> 
> I would assume that the hub was using sub-interface frame, with the key tied 
> to that specific interface/DLCI.
> 
> Let me see the config.
> 
> Good troubleshooting and debugging, by the way. Glad to see an educated 
> hypothesis based upon debugging. You are well on your way to the IE my friend.
> 
> Regards,
> Jay McMickle- CCIE #35355
> Sent from iJay
> 
> On May 3, 2012, at 6:38 AM, Rob Pool <[email protected]> wrote:
> 
> > George,
> > 
> > I don't believe more than one key can be active at a time on an interface. 
> > So it seems like the key to solving the issue is by creating addition 
> > interfaces whether they be tunnel, vitual-template or even subinterfaces. 
> > How did you configure gre tunnels or how would you configure pppofr without 
> > adding additional addressing which seems to be a requirement in most labs? 
> > Did you do ip unnumbered? 
> > 
> > Sent from my iPhone
> > 
> > On May 2, 2012, at 6:48 PM, George Leslie <[email protected]> 
> > wrote:
> > 
> >> 
> >> 
> >> 
> >> 
> >> Hi all,Came across an interesting little tidbit of info today while 
> >> playing around with EIGRP authentication on a frame hub and spoke network. 
> >> No doubt, you'll remember the IPE lab where you have a frame hub and 
> >> spoke, running OSPF, and you have to use different authentication keys for 
> >> each of the spokes?  Well, I tried doing the same with EIGRP 
> >> authentication, using key chains.  Hub had keys 1 and 2; spoke 1 had key 
> >> 1; spoke 2 had key 2.  All were valid keys: I had configured send and 
> >> accept lifetimes on ALL keys that started 00:00:00 1 jan 1993 and lasted 
> >> an infinite lifetime.  The "show key chain" command confirmed that ALL 
> >> keys were valid. The bahaviour I saw was that the neighbour relationship 
> >> between hub and spoke 1 was solid.  However, the neighbour relationship 
> >> between hub and spoke 2 continually flapped.  Hub would see it come up as 
> >> a valid neighbour, 180 hold time would expire, it would reset, come back 
> >> in again etc.  On spoke 2, you never saw the hub as a neig
 hbo
> > ur
> >> . Doing a bit of debug eigrp packet showed that the hub ONLY used key 1 
> >> and not key 2.  Hub would accept key 2 from spoke 2 but never send with 
> >> it.  Doesn't this defeat the point of having overlapping send and receive 
> >> lifetimes on the keys for key switchover?  The hub simply did not use the 
> >> second key, even although it was receiving and correctly authenticating 
> >> received packets with it! Firstly, does anyone know if there is some sort 
> >> of timeout here, when the hub reverts to using both keys?  I gave up 
> >> waiting (I spent about 10 minutes troubleshooting until I decided to try 
> >> another tack). My workaround in the end was to configure two GRE tunnels, 
> >> between each spoke and the hub, and move EIGRP away from the physical 
> >> interfaces and onto the tunnels, and use different key chains on the hub.  
> >> Worked a treat.  Suppose I could have used PPPoFR as well, but that would 
> >> have incurred more typing! Regards, George.                         
> >> _______________________________________________
> >> For more information regarding industry leading CCIE Lab training, please 
> >> visit www.ipexpert.com
> >> 
> >> Are you a CCNP or CCIE and looking for a job? Check out 
> >> www.PlatinumPlacement.com
> >> 
> >> http://onlinestudylist.com/mailman/listinfo/ccie_rs
> > _______________________________________________
> > For more information regarding industry leading CCIE Lab training, please 
> > visit www.ipexpert.com
> > 
> > Are you a CCNP or CCIE and looking for a job? Check out 
> > www.PlatinumPlacement.com
> > 
> > http://onlinestudylist.com/mailman/listinfo/ccie_rs
                                          
_______________________________________________
For more information regarding industry leading CCIE Lab training, please visit 
www.ipexpert.com

Are you a CCNP or CCIE and looking for a job? Check out 
www.PlatinumPlacement.com

http://onlinestudylist.com/mailman/listinfo/ccie_rs

Reply via email to