cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
33793
Views
16
Helpful
8
Replies

OSPF To many retransmissions

mahesh18
Level 6
Level 6

Hi all,

We have MAN conenction between two sites and both sites use OSPF  as Routing protocol.

I see on Link A and B  router following logs

000717: Mar  4 22:34:09.187 EST: %OSPF-5-ADJCHG: Process 3738, Nbr on GigabitEthernet4/4 from EXSTART to DOWN, Neighbor Down: Dead timer expired
000718: Mar  4 22:35:59.155 EST: %OSPF-5-ADJCHG: Process 3738, Nbr on GigabitEthernet4/4 from EXSTART to DOWN, Neighbor Down: Dead timer expired
000719: Mar  4 22:42:15.531 EST: %OSPF-5-ADJCHG: Process 3738, Nbr on GigabitEthernet4/4 from EXSTART to DOWN, Neighbor Down: Too many retransmissions
000720: Mar  4 22:43:15.531 EST: %OSPF-5-ADJCHG: Process 3738, Nbr on GigabitEthernet4/4 from DOWN to DOWN, Neighbor Down: Ignore timer expired
000721: Mar  4 22:45:25.620 EST: %OSPF-5-ADJCHG: Process 3738, Nbr  on GigabitEthernet4/4 from EXSTART to DOWN, Neighbor Down: Too many retransmissions
000722: Mar  4 22:46:25.620 EST: %OSPF-5-ADJCHG: Process 3738, Nbr on GigabitEthernet4/4 from DOWN to DOWN, Neighbor Down: Ignore timer expired
000723: Mar  4 22:49:09.148 EST: %OSPF-5-ADJCHG: Process 3738, Nbr  on GigabitEthernet4/4 from EXSTART to DOWN, Neighbor Down: Too many retransmissions
000724: Mar  4 22:50:09.148 EST: %OSPF-5-ADJCHG: Process 3738, Nbr  on GigabitEthernet4/4 from DOWN to DOWN, Neighbor Down: Ignore timer expired
000725: Mar  4 22:55:45.952 EST: %OSPF-5-ADJCHG: Process 3738, Nbr  on GigabitEthernet4/4 from EXSTART to DOWN, Neighbor Down: Too many retransmissions
000726: Mar  4 22:56:45.952 EST: %OSPF-5-ADJCHG: Process 3738, Nbr  on GigabitEthernet4/4 from DOWN to DOWN, Neighbor Down: Ignore timer expired
000727: Mar  4 23:00:40.885 EST: %OSPF-5-ADJCHG: Process 3738, Nbr  on GigabitEthernet4/4 from LOADING to FULL, Loading Done

Both sides show OSPF  is down due to Too many retransmissions

Can someone expalinme please what is Too many retransmissions  and how can we fix this?

Is this telco issue or something in cabling or config?

thanks

mahesh

2 Accepted Solutions

Accepted Solutions

Nagendra Kumar Nainar
Cisco Employee
Cisco Employee

Hi Mahesh,

I have seen this "too many retranmission" error message along with EXSTART to DOWN, when you have MTU issue (not MTU mismatch) between the neighbors.

In other words, when your MTU on the interface is say 2000 while for some reason, the router is not able to process pakt of size 2000. In this case, when OSPF is enabled, after initial DBD mesage exchange (with MTU value of 2000 advertised in DBD ), neighbor STATE will be moved to EXSTART and will start sending actual DBD with LSA headers.This DBD packet size will be based on interface MTU. In this case, it will be 2000 which the neighbor may not be able to process. So no ACK or LSA request will be received from neighbor. This will result in moving neighbor STATE from EXSTART to DOWN and same process will continue.

Simply try sending ping traffic to size of outgoing MTU in interface and DF bit set. That will show if you are hitting MTU issue.

HTH,

Nagendra

View solution in original post

8 Replies 8

Nagendra Kumar Nainar
Cisco Employee
Cisco Employee

Hi Mahesh,

I have seen this "too many retranmission" error message along with EXSTART to DOWN, when you have MTU issue (not MTU mismatch) between the neighbors.

In other words, when your MTU on the interface is say 2000 while for some reason, the router is not able to process pakt of size 2000. In this case, when OSPF is enabled, after initial DBD mesage exchange (with MTU value of 2000 advertised in DBD ), neighbor STATE will be moved to EXSTART and will start sending actual DBD with LSA headers.This DBD packet size will be based on interface MTU. In this case, it will be 2000 which the neighbor may not be able to process. So no ACK or LSA request will be received from neighbor. This will result in moving neighbor STATE from EXSTART to DOWN and same process will continue.

Simply try sending ping traffic to size of outgoing MTU in interface and DF bit set. That will show if you are hitting MTU issue.

HTH,

Nagendra

Hi,

Thanks for reply.

how can i do the ping test and check the size of mtu?

mahesh

Hi, You need to use the extended ping command. See below

http://www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a0080093f22.shtml

See also a handy MTU mismatch troubleshooting guide between two OSPF routers

http://www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a0080093f0d.shtml

Hi all,

Many thanks for the update.

regards

mahesh

@Nagendra 

You only discussed what's happening but you didn't mention a solution. 

How do we resolve this?

I'm also getting EXSTART to DOWN. I have 9100 ip mtu configured on both ends (one Cisco and one is a different vendor). 

When I configure ip mtu-ignore, the OSPF neighborship comes but, but I don't want to use mtu-ignore. 

Why can't one of the routers process a 9100 mtu packet?  

Hi @mfatty 

 

The simple solution to this problem is to use same MTU on both end routers running OSPF in the case of Cisco Routers. Although this becomes a bit complex when different vendor equipment is being used as each vendor may implement implement MTU differently. For example running 'ip mtu 1500' command configured on Cisco vs Juniper router interface will not have the same implications. This is because in Cisco 20byte IP header is included in 1500 but not in Juniper.

 

As suggested in the earlier reply the best way to verify this is to run an extended ping sweep.

 

Also even if both ends of the links are matching MTU but the underlying MTU is smaller this can cause retransmission if LSU, LSR or LSAck is being sent by either of the routers and is getting dropped by intermediary devices with smaller MTU. In such a scenario and Established OSPF neighborship will be torn down after 25 retransmissions. Please refer to the link below for further details.

 

https://www.cisco.com/c/en/us/support/docs/ip/open-shortest-path-first-ospf/116119-technote-ospf-mtu-00.html

https://www.cisco.com/c/en/us/support/docs/ip/open-shortest-path-first-ospf/119433-technote-ospf-00.html

There are other reasons the OSPF adjacency could fail with the log message "Too many retransmissions". For example, two switches connected with a Layer 3 Port Channel, which might be misconfigured at one end, such that the Port Channel comes up, but does not have all of the constituent switchports configured correctly.

Here is an example of what I mean:

 

SW01#show run inter port 1
Building configuration...

Current configuration : 86 bytes
!
interface Port-channel1
no switchport
ip address 192.168.x.1 255.255.255.252
end

SW01#show run inter gig0/1
Building configuration...

Current configuration : 162 bytes
!
interface GigabitEthernet0/1
no switchport
no ip address
negotiation auto
channel-group 1 mode on
end

SW01#show run inter gig0/2
Building configuration...

Current configuration : 107 bytes
!
interface GigabitEthernet0/2
negotiation auto
end

SW01#

SW01#
%OSPF-5-ADJCHG: Process 1, Nbr 192.168.x.254 on Port-channel1 from FULL to DOWN, Neighbor Down: Dead timer expired

=======================================

SW02#
%OSPF-5-ADJCHG: Process 1, Nbr 192.168.x.253 on Port-channel2 from DOWN to DOWN, Neighbor Down: Ignore timer expired
SW02#show run inter port 2
Building configuration...

Current configuration : 116 bytes
!
interface Port-channel2
no switchport
ip address 192.168.x.2 255.255.255.252
no logging event link-status
end

SW02#show run inter gig0/1
Building configuration...

Current configuration : 223 bytes
!
interface GigabitEthernet0/1
no switchport
no ip address
negotiation auto
channel-group 2 mode on
end

SW02#show run inter gig0/2
Building configuration...

Current configuration : 223 bytes
!
interface GigabitEthernet0/2
no switchport
no ip address
negotiation auto
channel-group 2 mode on
end

SW02#
%OSPF-5-ADJCHG: Process 1, Nbr 192.168.x.253 on Port-channel2 from EXSTART to DOWN, Neighbor Down: Too many retransmissions
SW02#

 

The resolution in this case is to ensure that the interface GigabitEthernet0/2 on SW01 has the correct configuration added to it.

 

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community:

Innovations in Cisco Full Stack Observability - A new webinar from Cisco