Cisco Support Community
cancel
Showing results for 
Search instead for 
Did you mean: 
Announcements

Welcome to Cisco Support Community. We would love to have your feedback.

For an introduction to the new site, click here. If you'd prefer to explore, try our test area to get started. And see here for current known issues.

New Member

LDP session flapping

Hi,

I have experienced problem on an ASR 1000 router with MPLS over DMVPN, whose LDP session is flapping periodically aproximatelly every 3 minutes.

The router is a DMVPN spoke with software 3.7S4. On central site, I have an ASR 1000 DMVPN hub router with 3.7S1 software

The spoke router-id i(tied to loopback0) is 10.100.249.24, hub router-id is 10.100.255.110.

We run OSPF on the tunnel interface and the ospf neighborship is up and running - no flaps, no problems on that side.

Here are the spoke periodical logs:

012742: Dec  2 11:53:41.518 MET: %LDP-5-NBRCHG: LDP Neighbor 10.100.255.110:0 (1) is UP

012743: Dec  2 11:56:29.734 MET: %LDP-5-NBRCHG: LDP Neighbor 10.100.255.110:0 (1) is DOWN (Received error notification from peer: KeepAlive expired)

012744: Dec  2 11:56:37.534 MET: %LDP-5-NBRCHG: LDP Neighbor 10.100.255.110:0 (1) is UP

012745: Dec  2 11:59:32.481 MET: %LDP-5-NBRCHG: LDP Neighbor 10.100.255.110:0 (1) is DOWN (Received error notification from peer: KeepAlive expired)

012746: Dec  2 11:59:36.265 MET: %LDP-5-NBRCHG: LDP Neighbor 10.100.255.110:0 (3) is UP

012747: Dec  2 12:02:34.577 MET: %LDP-5-NBRCHG: LDP Neighbor 10.100.255.110:0 (3) is DOWN (Received error notification from peer: KeepAlive expired)

012748: Dec  2 12:02:37.831 MET: %LDP-5-NBRCHG: LDP Neighbor 10.100.255.110:0 (1) is UP

012749: Dec  2 12:05:37.821 MET: %LDP-5-NBRCHG: LDP Neighbor 10.100.255.110:0 (1) is DOWN (Received error notification from peer: KeepAlive expired)

.... And so on.

The central router probably really doesn't receive the LDP keepalives:

0cc1-r1#sh mpls ldp nei 10.100.249.24                        

    Peer LDP Ident: 10.100.249.24:0; Local LDP Ident 10.100.255.110:0

        TCP connection: 10.100.249.24.646 - 10.100.255.110.28800

        State: Oper; Msgs sent/rcvd: 1352/2; Downstream

        Up time: 00:01:03

        LDP discovery sources:

          Tunnel20, Src IP addr: 10.100.237.23

The strange thing is, that there are more than 27 routers (mix of 7200, ASR 1000, ISR) connected to the central site, but just this one is having problems.  I have compared the configs and didn't find any difference. I tried to shut - no shut tunnel, no change in behaviour.

The strangest thing is, that the problem has dissapeared for some 3 weeks without any reason, but has reappeared after the DMVPN tunnel interface went down and up recently.

Has anyone ever experieced this as well?

What might be the cause?

Do you have any clue where to look for troubleshooting?

Thanks,

Alexander

Everyone's tags (6)
1 ACCEPTED SOLUTION

Accepted Solutions

LDP session flapping

Hi Alexander,

I had a similar problem with LDP, but not used DMVPN.

May be issue in MSS.

Can you capture traffic on the interface on the central and the spoke routers.

Thanks,

Alex

2 REPLIES

LDP session flapping

Hi Alexander,

I had a similar problem with LDP, but not used DMVPN.

May be issue in MSS.

Can you capture traffic on the interface on the central and the spoke routers.

Thanks,

Alex

New Member

LDP session flapping

To wrap up this issue, I have tuned the TCP MSS value on the outgoing interface, which has solved the problem.

Thanks a lot for pointing me to the solution.

Alexander

731
Views
0
Helpful
2
Replies