I have two ASBR router both injecting external default route into OSPF: ASBR1 and ASBR2.
ASBR1 redistribute default if installed as type1
ASBR2 redistribute default if installed as type2
The local default route cost on ASBR2 is higher then the cost received by ospf from ASBR1: so when everything is working ASBR2 has active default route the default received by ospf from the ASBR1.
When ASBR2 looses the ospf peer from which is receiving the default route injected by ASBR1 this happens on ASBR2:
-the 0.0.0.0 route listed in ospf routes tab gets "unknown" interface (STRANGE
-the local default route in routing table becomes active (OK)
When the lost ospf peer cames back the 0.0.0.0 route is not correctly recoverd and the local default route remains active instead of the default router from ASBR1.
If on the ASBR2 I restart the ospf instance all comes back as it was supposed to be.
If the ASBR2 has only one neighbor the problem does not happen, when the peer goes down the ospf stops running completely, when the pees goes back ospf start from scratch. But if ASBR2 has more than one peer the extenal routes in ospf tab always hangs until ospf is restared.
The same apply when the external routes injected by the two ASBRs are external subnet.
Mikrotik OSPF implementation bugs?
