Spanning tree problem

[sdcoyle80]

Hi guys,

Just wonder if someone can help me with this strange issue I've come across. 

I have a client who has a branch office and the two sites are linked with a 1Gbps dark fibre service, there is also a 1Gbps microwave service as a backup/failover. At both sites there is a Cisco 3560 switch which hold the fibre modules for each link. So its practically the same as having two direct links between two switches

I have set up spanning tree so that the dark fibre is the primary link and the microwave is the failover. This was working as expected. I recently noticed that the dark fibre connection was regularly dropping and reconnecting on the branch office switch. I put this down to auto-negotation of the interface so I configured 'speed nonegotiate' on the interface and the connection has been stable, however the spanning-tree configuration has changed.

Before nonegotiate was set on branch office switch

WAN-SW#sh spanning-tree vlan 172

VLAN0172
Spanning tree enabled protocol rstp
Root ID Priority 24748
Address aca0.1691.6c80
Cost 4
Port 25 (GigabitEthernet0/25)
Hello Time 2 sec Max Age 20 sec Forward Delay 15 sec

Bridge ID Priority 32940 (priority 32768 sys-id-ext 172)
Address ecc8.82f9.6480
Hello Time 2 sec Max Age 20 sec Forward Delay 15 sec
Aging Time 300 sec

Interface Role Sts Cost Prio.Nbr Type
------------------- ---- --- --------- -------- --------------------------------
Gi0/25 Root FWD 4 128.25 P2p
Gi0/27 Altn BLK 4 128.27 P2p

Gi0/25 is the dark fibre, Gi0/27 is the microwave

After nonegotiate is configured on Gi0/25

WAN-SW#sh spanning-tree vlan 172

VLAN0172
Spanning tree enabled protocol rstp
Root ID Priority 24748
Address aca0.1691.6c80
Cost 4
Port 27 (GigabitEthernet0/27)
Hello Time 2 sec Max Age 20 sec Forward Delay 15 sec

Bridge ID Priority 32940 (priority 32768 sys-id-ext 172)
Address ecc8.82f9.6480
Hello Time 2 sec Max Age 20 sec Forward Delay 15 sec
Aging Time 300 sec

Interface Role Sts Cost Prio.Nbr Type
------------------- ---- --- --------- -------- --------------------------------
Gi0/25 Desg FWD 4 128.25 P2p
Gi0/27 Root FWD 4 128.27 P2p

The microwave interface is now the root port! And the dark fibre is a designated port, its not even blocked

I have tried configuring a higher cost on gi0/27 and a lowest port priority on gi0/25 but it still continues to show gi0/27 as the root port. If I remove the nonegotiate configuration on gi0/25 everything goes back to how it was.

Does disabling auto-negotiate "break" spanning-tree? 

Thanks in advance!

[Rolf Fischer]

Hi,

an explanation for the STP roles could be that Gi0/25 has stopped receiving BPDUs from the root switch at the other site when you disabled auto-negotiation (for whatever reasons).

Could you check if BPDU-receive counters are increasing or not? What about non-STP traffic? Do you see the other switch as a CDP neighbor on Gi0/25, and are the RX-counters on that interface still increasing?

HTH

Rolf

[sdcoyle80]

Hi Rolf,

Thanks for the reply

I did notice that there are no BPDUs received on Gi0/25 once its set to nonegotiate. There is only 1 CDP neighbor listed, on port Gi0/27. The RX-counters on Gi0/25 arent increasing either.

I'd like to shut the Gi0/27 interface to see if Gi0/25 even comes up as the new root port but I don't want to try that remotely.

Steven

[paul driver]

Hello

Seems strange that disabling auto-negotiate would incur this stp change unless disabling this on a gb interface disables the interface

Sh int X/X status

Can I ask when you changed the port-priority where did you make that change?

it needs to be upstream on the attaching switch taking  the port identifier on the received port for the correct priority cost.

sh spanning-tree interface X/X detail

res

Paul

[sdcoyle80]

Hi Paul,

Thanks for the reply

The port priority was set on the head office switch

Here is the output of those commands

WAN-SW#sh int gi0/25 status

Port Name Status Vlan Duplex Speed Type
Gi0/25 Telstra Fibre connected 172 full 1000 1000BaseSX SFP

WAN-SW#sh spanning-tree int gi0/25 detail
Port 25 (GigabitEthernet0/25) of VLAN0172 is designated forwarding
Port path cost 4, Port priority 128, Port Identifier 128.25.
Designated root has priority 24748, address aca0.1691.6c80
Designated bridge has priority 32940, address ecc8.82f9.6480
Designated port id is 128.25, designated path cost 4
Timers: message age 0, forward delay 0, hold 0
Number of transitions to forwarding state: 1
Link type is point-to-point by default
BPDU: sent 3341, received 0

Steven

[Rolf Fischer]

Hi Steven,

thanks for the additional information. Seems to me rather a link problem than a STP issue, because neither control traffic nor production traffic is received on Gi0/25. Maybe you should talk with your service provider:

http://herdingpackets.net/2013/03/21/disabling-gigabit-link-negotiation-on-fiber-interfaces/

HTH

Rolf

[sdcoyle80]

Hi Rolf,

Yes I'm starting to think its a problem with the provider. It wouldn't be the first time we've had problems with the fibre connection!

Thanks for link, I actually saw that page when I was doing my initial troubleshooting.

Steven

[sbhadrav@cisco.com]

Packet corruption can also lead to the same kind of failure. If a link has a high rate of physical errors, you can lose a certain number of consecutive BPDUs. This loss can lead a blocking port to transition to forwarding state. You do not see this case very often because STP default parameters are very conservative. The blocking port needs to miss BPDUs for 50 seconds before the transition to forwarding. The successful transmission of a single BPDU breaks the loop. This case commonly occurs with the careless adjustment of STP parameters. An example of an adjustment is max-age reduction.

Duplex mismatch, bad cables, or incorrect cable length can cause packet corruption. Refer to the document Troubleshooting Switch Port and Interface Problems for an explanation of CatOS and Cisco IOS Software error counter output.