From: Roger Wang (rwang@xxxxxxxxxxx)
Date: Sun May 28 2000 - 18:32:49 GMT-3
Hi, all,
I have a policy applied on s1 of r3, like so:
======================================
!
interface Serial1
ip address 10.34.1.1 255.255.0.0
no ip directed-broadcast
ip ospf interface-retry 0
ip policy route-map lab1
clockrate 2000000
!
======================================
r3#sh ip pol
Interface Route map
local lab1
Serial1 lab1
r3#sh route-m
route-map lab1, permit, sequence 10
Match clauses:
ip next-hop (access-lists): 1
Set clauses:
ip next-hop 10.10.1.1
Policy routing matches: 10043 packets, 658889 bytes
r3#sh access-l 1
Standard IP access list 1
permit 10.10.1.2
permit 10.10.1.5
======================================
At the other end of the s1 interface is s0 of r4. I also configured a
virtual link between the two routers. The problem is, if I reload r4, r3
and r4 then can't become adjacent unless I remove route-map lab1 from r3's
s1.
debug ip os adj on r3:
--------------------------------------
05:36:53: OSPF: Send DBD to 10.44.2.1 on OSPF_VL0 seq 0x14D5 opt 0x22 flag
0x7 len 32
05:36:58: OSPF: Retransmitting DBD to 10.44.2.1 on OSPF_VL0
--------------------------------------
debug ip os adj on r4:
--------------------------------------
00:03:45: OSPF: Rcv DBD from 10.34.1.1 on OSPF_VL0 seq 0x14D5 opt 0x22 flag
0x7 len 32 state EXSTART
00:03:45: OSPF: First DBD and we are not SLAVE
00:03:49: OSPF: Retransmitting DBD to 10.34.1.1 on OSPF_VL0
00:03:49: OSPF: Send DBD to 10.34.1.1 on OSPF_VL0 seq 0x118 opt 0x22 flag
0x7 len 32
00:03:50: OSPF: Rcv DBD from 10.34.1.1 on OSPF_VL0 seq 0x14D5 opt 0x22 flag
0x7 len 32 state EXSTART
00:03:50: OSPF: First DBD and we are not SLAVE
--------------------------------------
after I removed route-map from r3's s1 (below is from r3):
--------------------------------------
05:37:08: OSPF: Rcv hello from 10.44.2.1 area 0 from OSPF_VL0 10.34.1.2
05:37:08: OSPF: End of hello processing
05:37:08: OSPF: Retransmitting DBD to 10.44.2.1 on OSPF_VL0
05:37:08: OSPF: Send DBD to 10.44.2.1 on OSPF_VL0 seq 0x14D5 opt 0x22 flag
0x7 len 32
05:37:08: OSPF: Rcv hello from 10.44.2.1 area 3 from Serial1 10.34.1.2
05:37:09: OSPF: End of hello processing
05:37:11: OSPF: Rcv DBD from 10.44.2.1 on OSPF_VL0 seq 0x118 opt 0x22 flag
0x7 len 32 mtu 0 state EXSTART
05:37:11: OSPF: NBR Negotiation Done. We are the SLAVE
05:37:11: OSPF: Send DBD to 10.44.2.1 on OSPF_VL0 seq 0x118 opt 0x22 flag
0x2 len 312
05:37:12: OSPF: Rcv DBD from 10.44.2.1 on OSPF_VL0 seq 0x119 opt 0x22 flag
0x3 len 112 mtu 0 state EXCHANGE
05:37:12: OSPF: Send DBD to 10.44.2.1 on OSPF_VL0 seq 0x119 opt 0x22 flag
0x0 len 32
05:37:12: OSPF: Rcv DBD from 10.44.2.1 on OSPF_VL0 seq 0x11A opt 0x22 flag
0x1 len 32 mtu 0 state EXCHANGE
05:37:12: OSPF: Exchange Done with 10.44.2.1 on OSPF_VL0
05:37:12: OSPF: Synchronized with 10.44.2.1 on OSPF_VL0, state FULL
05:37:12: OSPF: Send DBD to 10.44.2.1 on OSPF_VL0 seq 0x11A opt 0x22 flag
0x0 len 32
05:37:12: OSPF: Build router LSA for area 0, router ID 10.34.1.1, seq
0x80000
-----------------------------------
then r3 and r4 became adjacent. And if I apply the route-map back on s1 on
r3, the neighbor state still stays at "FULL".
It looks like r3 and r4 had problems deciding who's the Master/Slave during
the exstart state. The neighbor with the highest interface address should
become the master. Yet it didn't look like my route-map can affect this
decision...
Any ideas?
Thanks,
Rog
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 08:23:32 GMT-3