OSPF STUCK IN EXSTART – MTU MISMATCH
This post will detail on OSPF problem condition of “Stuck in EXSTART” state due to MTU mismatch between OSPF neighbors
The above diagram depicts 2 Routers R1 and R2 connected to each via Ethernet interface fa0/0 over 192.168.12.0/24 subnet. Interface fa0/0 of R1 is assigned IP 192.168.12.1 while R2’s Fa0/0 is assigned 192.168.12.2.Routers are running OSPF (Area 0).The routers form full adjacency with R2 being the DR (Designated Router).
As depicted in the above diagram, for understanding the effect of MTU on OSPF neighborship, we change the MTU size on Interface Fa0/0 of R2 from 1500 to 1499 bytes. The command on interface level is “ip mtu 1499”. Now, we enable “debug ip ospf adj” command on each of R1 and R2 to have a detailed view of OSPF activities while OSPF adjacency between R1 and R2 is in process.
R1 debug output shows that it is receiving smaller MTU from R2 while R2 debug output shows it is receiving larger MTUfrom R1.
In this case, R2 will drop the OSPF packet received from R1 and will stay stuck in “EXSTART” state since it cannot accept OSPF packet with size larger MTU than its interface MTU size. However, R1 accepts the OSPF packet and moves to exchange state.
As shown in above diagram, in order to resolve the issue of MTU mismatch and therefore stuck in “EXSTART” state, we can
(1) Either change MTU size on Interface of R1 to 1499 Bytes
(2) Tell R1 to ignore MTU check during EXSTART stage.
Here we will use option 2.
For performing this change, we go to interface fa0/0 of R1 and add “ip ospf mtu-ignore” command.
Once done, R2 will bypass the mtu check process in EXSTART stage and further move to full adjacency. The “show ip ospf“output shows both the Routers having full OSPF state.