UDLD problem Link detection

[sebastien3]

Hello,

I want to use UDLD in this configuration but I notice a big problem if I cut the cable between the two media converters in the center of the image :

UDLD detects the outage and shuts down the ports on switches (Te1/16). Then after 30s (recovery udld) it reactivates the port Te1/16 while the connection between the two media converters is cut...

*Oct 16 23:23:13.210: %UDLD-4-UDLD_PORT_DISABLED: UDLD disabled interface Te1/16, aggressive mode failure detected
*Oct 16 23:23:13.210: %PM-4-ERR_DISABLE: udld error detected on Te1/16, putting Te1/16 in err-disable state
*Oct 16 23:23:14.210: %LINEPROTO-5-UPDOWN: Line protocol on Interface TenGigabitEthernet1/16, changed state to down
*Oct 16 23:23:15.210: %LINK-3-UPDOWN: Interface Vlan1, changed state to down
*Oct 16 23:23:15.211: %LINK-3-UPDOWN: Interface TenGigabitEthernet1/16, changed state to down
*Oct 16 23:23:16.210: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan1, changed state to down

*Oct 16 23:23:43.207: %PM-4-ERR_RECOVER: Attempting to recover from udld err-disable state on Te1/16
*Oct 16 23:23:46.075: %LINK-3-UPDOWN: Interface TenGigabitEthernet1/16, changed state to up
*Oct 16 23:23:47.075: %LINEPROTO-5-UPDOWN: Line protocol on Interface TenGigabitEthernet1/16, changed state to up
*Oct 16 23:24:16.075: %LINK-3-UPDOWN: Interface Vlan1, changed state to up
*Oct 16 23:24:17.075: %LINEPROTO-5-UPDOWN: Line protocol on Interface Vlan1, changed state to up

Switch#sh udld neighbors
Port     Device Name   Device ID     Port ID    Neighbor State
----     -----------   ---------     -------    --------------

Switch#sh spanning-tree vlan 1

VLAN0001
  Spanning tree enabled protocol ieee
  Root ID    Priority    32769
             Address     286f.7f24.ef00
             This bridge is the root
             Hello Time   2 sec  Max Age 20 sec  Forward Delay 15 sec

  Bridge ID  Priority    32769  (priority 32768 sys-id-ext 1)
             Address     286f.7f24.ef00
             Hello Time   2 sec  Max Age 20 sec  Forward Delay 15 sec
             Aging Time  15  sec

Interface           Role Sts Cost      Prio.Nbr Type
------------------- ---- --- --------- -------- --------------------------------
Te1/16              Desg FWD 4         128.16   P2p

!
errdisable recovery cause udld
errdisable recovery interval 30
!
interface TenGigabitEthernet1/16
 udld port aggressive
 udld fast-hello 200
!

The problem is that traffic will fail ! Is there a function to force the port to stay DOWN when the neighbor is not reachable ?

Use loopguard on port Te1/16 to avoid having traffic on this interface ?

Thanks.

[marce1000]

 - Check port configuration on Te1/16 , are you using any form of auto recovery (commands) ?

 M.

[sebastien3]

@marce1000The config of port Te1/16 is in the last post.

!
errdisable recovery cause udld
errdisable recovery interval 30
!
interface TenGigabitEthernet1/16
udld port aggressive
udld fast-hello 200
!

[marce1000]

 - Remove the errdisable recovery commands if you  don't want automatic recovery of an udld failure ,

 M.

[sebastien3]

I would like the port to automatically go UP but not let traffic pass if its neighbor does not respond... Possible to do this ?

[MHM Cisco World]

It not easy' the admin must handle some time.

But for me the best solution is 

1- remove recovery 

2- config eem to no shut ther errdisable port after wait for example 10 min.

3- admin cab no shut the port manually in time between eem detect errdisable and action no shut

That only way I see.

[sebastien3]

Hmm I see but I still have a doubt about the solution. Imagine the cable between the two media converters is cut... I don't know what the recovery time will be.

During downtime the port must remain DOWN so as not to have traffic sent on this interface.

[marce1000]

   >During downtime the port must remain DOWN so as not to have traffic sent on this interface.
  That's why you  don't want  any form of   errdisable recovery commands on the port configuration, 

 M.

[sebastien3]

I think we don't understand each other.

I'm looking for a solution but it doesn't seem to exist ! If a cut occurs on the fiber or a person accidentally unplugs the fiber, UDLD detects the cut instantly and shuts the port, this part is OK.

I want the port to reactivate (no shut) and remain in Learned or Blocked mode in order to check if UDLD can reach its neighbor. Once UDLD is OK again, the port can go Forward.

[MHM Cisco World]

Stp is l2 

Udld is l1 

The time udld detects neighbor down is so so fast that make stp not recover from blk status and hence no traffic pass.

So you ask about errdisable' make it less than time stp need to go to forward status.

Note:- Errdisable recovery is effect all port even that not down by udld but down from other protocol.

[marce1000]

   >...I want the port to reactivate (no shut) and remain in Learned or Blocked mode in order to check if UDLD can reach its neighbor. Once UDLD is OK again, the port can go Forward.
            - That can't be done, UDLD sits at layer1 it doesn't know how 'to reach its neighbor'

 M.

[sebastien3]

Yes I agree, that's why I said the solution doesn't seem to exist...

[marce1000]

 - In essence that being a good think , because it sits as layer1 or cabling issue, physical work or inspection, or using devices like fluke to analyze the fiber is more appropriate, 

  M.