cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
2194
Views
0
Helpful
6
Replies

OSPF timers tuning

sergey.sh
Level 1
Level 1

We have 3 separate 3850 connected to 6807 with fiber

and timers on interfaces are

 ip ospf dead-interval 3
 ip ospf hello-interval 1

 

3850-1  have couple times per day

Port-channel1 from FULL to DOWN, Neighbor Down: Dead timer expired
Port-channel1 from LOADING to FULL, Loading Done

3850-2  have couple times per week 

Port-channel1 from FULL to DOWN, Neighbor Down: Dead timer expired
Port-channel1 from LOADING to FULL, Loading Done

3850-3  no logs OSPF is stable

 

no errors on all switche  interfaces

fiber were cleaned

3850-1  switch was replaced, but still have issue.

 

Cisco support recommend use default timers to resolve issue. Any idea? Is it timer issue? Any way running OSPF?

 

Thank you.

 

6 Replies 6

Zach S
Level 1
Level 1

If it's not causing any issues, you probably don't need to worry. The purpose of dead timers is to consider the link no longer valid after the timer has expired, due to a lack of OSPF hellos.

If you want, you can increase your dead timer a little bit, to 5 seconds, and the link flapping or going down for 3 seconds won't give you the error, but that won't fix your link flapping. Be aware that dead timers must to match between OSPF neighbors or the neighborship breaks, so if you change on one, change it on the rest.

Your issue isn't with OSPF itself, it's caused by those links dropping for more than 3 seconds at any given time. You can try swapping to different fibers to see if it resolves the issue.

I cleaned fiber, replace SFP's on both ends. No errors on interfaces. CPU 5-10% .

Joseph W. Doherty
Hall of Fame
Hall of Fame

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

If the links aren't actually dropping, you may just be losing the hello packets within a burst of other traffic.  Do the ports indicate any drops?

What kind of link is this?  I assume Ethernet, as you mention using 3850s, but will the link indicate a link down to both sides?  If it does, default times should be fine.  Only if the link doesn't indicate link down to both sides do you need to depend on the routing protocol's loss of neighbor timers.

If that's what's happening, you either need to increase the overall dead time interval (this to try to get some hellos outside a burst event), or you need to increase the number of hello packets within the same time interval (this to try to have some hello get through by having more [in your case fast hellos, or, if supported on the 3850, BPD]), or you use QoS to insure hello packets are protected against other traffic during burst events.

no errors on interfaces

no traffic on switches, just connected to Core.

fiber is  multy mode and single mode (short distance just other building)

Core switches CPU ~20%

3850 does not support BFD

Hello

on the spoke ospf peers

enable debug

 

Debutg ip ospf hello

debug ip ospf adj

 

and post the output

 

Res

Paul

 


Please rate and mark as an accepted solution if you have found any of the information provided useful.
This then could assist others on these forums to find a valuable answer and broadens the community’s global network.

Kind Regards
Paul

3850-1#
000222: Oct  7 20:44:53.933 PDT: %OSPF-5-ADJCHG: Process 20, Nbr 10.1.1.254 on Port-channel2 from FULL to DOWN, Neighbor Down: Dead timer expired
000223: Oct  7 20:44:53.936 PDT: %OSPF-5-ADJCHG: Process 20, Nbr 10.1.1.254 on Port-channel2 from DOWN to INIT, Received Hello
000224: Oct  7 20:44:53.936 PDT: %OSPF-5-ADJCHG: Process 20, Nbr 10.1.1.254 on Port-channel2 from INIT to 2WAY, 2-Way Received
000225: Oct  7 20:44:53.936 PDT: %OSPF-5-ADJCHG: Process 20, Nbr 10.1.1.254 on Port-channel2 from 2WAY to EXSTART, AdjOK?
000226: Oct  7 20:44:53.945 PDT: %OSPF-5-ADJCHG: Process 20, Nbr 10.1.1.254 on Port-channel2 from EXSTART to EXCHANGE, Negotiation Done
000227: Oct  7 20:44:53.962 PDT: %OSPF-5-ADJCHG: Process 20, Nbr 10.1.1.254 on Port-channel2 from EXCHANGE to LOADING, Exchange Done
000228: Oct  7 20:44:53.962 PDT: %OSPF-5-ADJCHG: Process 20, Nbr 10.1.1.254 on Port-channel2 from LOADING to FULL, Loading Done
Review Cisco Networking for a $25 gift card