We have two 4500 routers which is R1 and R2. The R1 and R2 has two interconnect which are primary and secondary link for redundant. For the primary interconnect, we use bandwidth shaper in between R1 and R2. For secondary link its just a direct connection between R1 and R2.
R1 interface configured with 1500 MTU size and R2 with 1520 MTU.
OSPF is running in between these two routers.
R1 interface configured "ip ospf mtu-ignore" but R2 none.
The situation is that, the OSPF adjacency no issue on primary link but when shifted to secondary link, the issue started wherby it built adjancy with FULL and after 1 minutes if shows DOWN. After 30 minutes, its back to FULL and DOWN.
Can i know what was the reason behind this?
Why the link which have bandwidth shaper in between does not have problem?
Does not anything to do with control/data plane?
OSFP is like "a litthe girl"....if everything<< timers, mtu,auth,area,subnet,mask>> are not in sync...full adj will NOT come up.
So, you have to issue mtu-ignore on the R2.
But that is the answer which i am scratching my head now. Both R1/R2 interface in primary and secondary config are same. But why it works good in primary link but not in secondary?
The only difference between these primary and secondary is that primary link have bandwidth shaper in between R1 and R2 but secondary link is just a direct connection between R1 and R2
If you take the shaper out of the middle does everything work fine? I would assume so. I wonder if the ospf traffic is being shaped, therefore causing an issue. If you do a "show ip ospf neigh" while you are waiting for the adjacency to form, does it stay in EXSTART for a long time?
Also, can you do a "debug ip ospf adjacency" on both ends?
There is no problem on Primary link which has the shaper. The problem is on secondary link without the shaper (direct connection with R1 and R2).
Thanks for the clarification. Can you do
"debug ip ospf adjacency"
on both the routers in question?