Hi,
In DMVPN Phase 2, EIGRP does not change the next hop to itself, so that the
spoke router has RIB information which points directly to the other spoke.
Based on RIB, CEF creates FIB and since there is no L2 adjacency to that
next hop it marks the entry Glean.
Now, when the data packet goes out, CEF cannot forward the packet and
consults NHRP table for a solution. There is no solution at this point yet,
so NHRP tells the CEF to use Hub IP address for the time being. Also, NHRP
triggers the process to get that information, so sends out NHRP Resolution
packet to the Hub, the Hub forwards it to the remote spoke, the remote
spoke sets up the direct tunnel back to the originating spoke and sends
down NHRP Reply with all required information. Once that information is in
NHRP table, CEF is happy and has L2 adjacency based on NHRP info.
Regards,
-- Piotr Matusiak CCIE #19860 (R&S, Security), CCSI #33705 Technical Instructor website: www.MicronicsTraining.com blog: www.ccie1.com If you can't explain it simply, you don't understand it well enough - Albert Einstein 2012/9/28 jeremy co <jeremy.cool14_at_gmail.com> > HI, > > I was studying DMVPN and came across CEF glean adjacency in phase 2. > > spoke2 get the routing update with next hop unchanged from router 1. on > Petr INE article there is sentence "this is the key to trigger CEF next-hop > resolution." > > > My question is exactly how CEF knows that it's not a normal EIGRP update > and it needs to clear a glean adj not invalid one. > > > Thanks! > > Jeremy > > > 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.netReceived on Thu Oct 04 2012 - 23:03:01 ART
This archive was generated by hypermail 2.2.0 : Thu Nov 01 2012 - 10:53:33 ART