Re: assist with eigrp pe-ce showing zero successor for a

Actually, I owe you all an apology. You can lie that it's NOT an EIGRP
route, but not artificially create one (even though I remember doing
that).

The "secret" is in EIGRP extended communities:

R2#sh ip bgp vpnv4 vrf a 10.0.0.5
BGP routing table entry for 24:24:10.0.0.5/32, version 8
Paths: (1 available, best #1, table a)
  Advertised to update-groups:
        1
  Local
    10.2.5.5 from 0.0.0.0 (192.168.0.2)
      Origin incomplete, metric 2297856, localpref 100, weight 32768,
valid, sourced, best
      Extended Community: RT:24:24
        Cost:pre-bestpath:128:2297856 (default-2145185791) 0x8800:32768:0
        0x8801:25:640000 0x8802:65281:1657856 0x8803:65281:1500
      mpls labels in/out 18/nolabel

Take a look at all those 0x880x ones. They define EIGRP metric and
metric type. You can delete them using "extcomm-list delete", but not
manually set them, which was what I was alluding at. Sorry :-)

--
Marko Milivojevic - CCIE #18427 (SP R&S)
Senior CCIE Instructor - IPexpert
On Mon, Feb 20, 2012 at 01:07, Marko Milivojevic <markom_at_ipexpert.com> wrote:
> Actually no, it will be on the one that runs OSPF. You need to lie to
> the other one how these routes came from EIGRP and not OSPF... :-)
>
> --
> Marko Milivojevic - CCIE #18427 (SP R&S)
> Senior CCIE Instructor - IPexpert
>
> On Sun, Feb 19, 2012 at 23:29, Rich Collins <nilsi2002_at_gmail.com> wrote:
>> I've been looking at your exercise and could it till now but am
>> curious to hear the answer on this one.
>> I am assuming that this trick is to be done on the PE running the IGP EIGRP.
>>
>> -Rich
>>
>> On Sun, Feb 19, 2012 at 3:15 AM, Marko Milivojevic <markom_at_ipexpert.com> wrote:
>>> You got it all -- well done. Now, as an exercise, try tricking your
>>> EIGRP on the receiving end to accept that OSPF-injected route as an
>>> internal one ;-)
>>>
>>> --
>>> Marko Milivojevic - CCIE #18427 (SP R&S)
>>> Senior CCIE Instructor - IPexpert
Blogs and organic groups at http://www.ccie.net