OSPF Adjacency issue

[_SiD_]

Good day team!

1 unit from stack 3850 (03.06.00E) went to reboot...after that, OSPF adj took 4 hours to complete....(like it were waiting for default ARP timeout) does anyone has seen this before?

attached is scheme and logs

[Leo Laohoo]

IOS-XE version 3.6.0?  Very brave to use a version with "0" at the end.  Very, very brave.  

[_SiD_]

i noted that to and planned to upgrade to 03.06.10

[_SiD_]

found on some topic - almost the same issue, but with no solution.

"you are using multicast, and multiaccesss environment, one of the routers keep sending prune message, which effect the ospf relation."

[MHM Cisco World]

there is L2 security called sticky ARP, this make ARP table refuse the modification of IP-MAC and hence the connection is lost, 
during reboot I think the Port MAC is change but the IP use between both OSPF still same this make traffic drop. 

check the sticky ARP and if you want disable it monitor OSPF and if it solve issue re enable it.