Protected ports blocking ospf hello, but Virtual Link still

From: Ladee Geek <ladeegeek_at_gmail.com>
Date: Mon, 24 May 2010 10:11:35 -0400

     (area 0) ---- R1*---------*(area 10)--------R4----(area 104)
                             |
                            R6

Hi All -

in this sutuation R1, R6 and R4 are all on the same switch and in the same
area 10. They are all configured with point-to-multipoint. The switchports
for R1 and R4 are configured with "switchport protected." Since area 104
lys on the other side of R4 a virutal link is required through area 10.

A debug on R1 showed that it was only getting hellos from R6.

R1#
*May 24 14:04:36.612: OSPF: Rcv hello from 172.16.106.1 area 10 from
FastEthernet0/1 172.16.10.6
*May 24 14:04:36.612: OSPF: End of hello processing
R1#
*May 24 14:04:56.484: OSPF: Send hello to 224.0.0.5 area 10 on
FastEthernet0/1 from 172.16.10.1
R1#
*May 24 14:05:05.736: OSPF: Rcv hello from 172.16.106.1 area 10 from
FastEthernet0/1 172.16.10.6
*May 24 14:05:05.736: OSPF: End of hello processing

When I did a show ip os neigh on R1 and R4 they both only have R6 as
neighbors.

R1(config-router)#do s ip os neig
Neighbor ID Pri State Dead Time Address Interface
172.16.102.1 0 FULL/ - 00:00:38 172.16.21.2
FastEthernet0/0
172.16.104.1 0 FULL/ - - 172.16.10.4 *OSPF_VL0*
172.16.106.1 0 FULL/ - 00:01:34 172.16.10.6
FastEthernet0/1

In the mean time both R4 and R1 have virtual links configured directly to
each other and the virtual link's adjacency state is full?

So - how is it that? Is the virtual link being routed via R6. It would
have to be.

R1#sho ip os virtual
Virtual Link OSPF_VL0 to router 172.16.104.1 is up
  Run as demand circuit
  DoNotAge LSA allowed.
  Transit area 10, via interface FastEthernet0/1, Cost of using 2
  Transmit Delay is 1 sec, State POINT_TO_POINT,
  Timer intervals configured, Hello 10, Dead 40, Wait 40, Retransmit 5
    Hello due in 00:00:08
    Adjacency State FULL (Hello suppressed)
    Index 2/3, retransmission queue length 0, number of retransmission 0
    First 0x0(0)/0x0(0) Next 0x0(0)/0x0(0)
    Last retransmission scan length is 0, maximum is 0
    Last retransmission scan time is 0 msec, maximum is 0 msec
  Message digest authentication enabled
    Youngest key id is 1

So in the end I have a question about the best practice for virtual links.
On a multiaccess network as in area 10, if there are no other elements like
ports are protected and we are dealing with a plain old vanilla area 10 and
needs to be a transit for a VL where should the end points be?

Should the VL go from R1-R6 and R6 to R4? or just R1-R4?

I've seen scenarios where the vl went from r1-r6 and r6-r4. But I can only
think that that would be needed if R6 was an ABR as well.

Thoughts?

r/
LG

Blogs and organic groups at http://www.ccie.net
Received on Mon May 24 2010 - 10:11:35 ART

This archive was generated by hypermail 2.2.0 : Tue Jun 01 2010 - 07:09:53 ART