06-26-2008 08:07 PM - edited 03-05-2019 11:51 PM
Dear Folks,
We had old 3500 Series Switches which we have replaced last week with 2960 48 Port ehternet and 2 Gigabit Port with SFP Module.
Now I am facing the problem : The Port of CORE SW to which my one of 2960 SW is connected got in to errordisable. UDLD is enable on CORe as well UDLD enable aggressively on 2960 Gi0/1.
It's understandable that port was being errordisable if unidirectional link is there but the same link is working previously with 3500 Series SW?
Than IS there any other reason for errordisable or something need to change on 2960 SW.
2960 SW basic Config :
config terminal
username -------- password .........
enable secret ...........
enable password .......
service password-encryption
spanning-tree portfast bpduguard
int range fa0/1 - 48
spanning-tree portfast
exit
vtp mode transparent
vtp mode client
vtp domain RILHZ
clock timezone GMT 5 30
ntp server 10.2.202.52
int range gi0/1 - 2
switchport mode trunk
udld port aggressive
int vlan 1
ip address x.x.x.x y.y.y.y
exit
line vty 0 4
password ..........
no login
login local
exit
My other end is 6506 L3 Switch with Cat OS and in it we have configured udld on all port i.e.
#udld
set udld enable
#module 1 : 2-port 1000BaseX Supervisor
set port trap 1/1-2 enable
set udld enable 1/1-2
set trunk 1/1 on dot1q 1-1005,1025-4094
set trunk 1/2 on dot1q 1-1005,1025-4094
set port channel 1/1-2 mode on
!
#module 2 : 2-port 1000BaseX Supervisor
set port trap 2/1-2 enable
set trunk 2/1 on dot1q 1-1005,1025-4094
set trunk 2/2 on dot1q 1-1005,1025-4094
set port channel 2/1-2 mode on
!
#module 3 : 16-port 1000BaseX Ethernet
set port trap 3/1-16 enable
set trunk 3/1 on dot1q 1-1005,1025-4094
set trunk 3/2 on dot1q 1-1005,1025-4094
set trunk 3/3 on dot1q 1-1005,1025-4094
set trunk 3/4 on dot1q 1-1005,1025-4094
set trunk 3/5 on dot1q 1-1005,1025-4094
set trunk 3/6 on dot1q 1-1005,1025-4094
set trunk 3/7 on dot1q 1-1005,1025-4094
set trunk 3/8 on dot1q 1-1005,1025-4094
set trunk 3/9 on dot1q 1-1005,1025-4094
set trunk 3/10 on dot1q 1-1005,1025-4094
set trunk 3/11 on dot1q 1-1005,1025-4094
set trunk 3/12 on dot1q 1-1005,1025-4094
set trunk 3/13 on dot1q 1-1005,1025-4094
set trunk 3/14 on dot1q 1-1005,1025-4094
set trunk 3/15 on dot1q 1-1005,1025-4094
set trunk 3/16 on dot1q 1-1005,1025-4094
set spantree portcost 3/1 3
!
#module 4 empty
!
#module 5 : 16-port 1000BaseX Ethernet
set port trap 5/1-16 enable
set udld enable 5/3-4,5/6-7
set trunk 5/1 on dot1q 1-1005,1025-4094
set trunk 5/2 on dot1q 1-1005,1025-4094
set trunk 5/3 on dot1q 1-1005,1025-4094
set trunk 5/4 on dot1q 1-1005,1025-4094
set trunk 5/15 on dot1q 1-1005,1025-4094
!
The Above is configuration for UDLD on Core SW.
On 2960 UDLD is enabled only on gi0/1 and gi0/2 Port in aggressive mode.
Portfast is enable on all port fa0/1 - 48
portfastbpdugurd is enable default mode i.e globally.
FO cable is same as before when old 3500 series SW is working. But after migrating sw i.e 2960 this proble occurs.
And avg response of all 2960 sw is also slow now.
I am attaching the basic Topology dia which can help you to understand the problem properly.
Pls do needful as soon as possible.................
Dipesh
06-26-2008 11:30 PM
Dear Sir,
One more thing I found that Fo cable is Multimode and at 2960 side Patch cord is LC-ST Single mode with Single Mode SFP module.........
At core side Single mode GBIC with Multimode SC-ST Patch Cord.
This might be the Problem?
But this is working on 3500 Series with Single mode Patch cord and single mode GBIC.
Here is Packet ping Drop result for affected sw :
Ping statistics for x.x.x.x:
Packets: Sent = 11252, Received = 11250, Lost = 2 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 0ms, Maximum = 16ms, Average = 0ms
Control-C
Ping statistics for x.x.x.x:
Packets: Sent = 12191, Received = 11859, Lost = 332 (2% loss),
Approximate round trip times in milli-seconds:
Minimum = 0ms, Maximum = 31ms, Average = 0ms
Control-C
Ping statistics for x.x.x.x:
Packets: Sent = 11924, Received = 11580, Lost = 344 (2% loss),
Approximate round trip times in milli-seconds:
Minimum = 0ms, Maximum = 16ms, Average = 0ms
Control-C
All users are affected due to this.
06-26-2008 11:47 PM
Hi Dipesh
Since you use the encapsulation on the 6506 as dot1q, you would need to this on the 2960 switch as well .... and also it looks like you have channeling turned on for ports 1/1-2 on the 6506. If these goes to the 2960, then you need the following commands on the 2960 ...
2960 Switch
-----------
int port-channel 1
switchport trunk encap dot1q
switchpport mode trunk
!
!
int range gi0/1- 2
switchport trunk encap dot1q
channel-group 1 mode on
no shut
* Also note that the fiber has to be end-to-end multi mode or single mode.
Cheers,
- sn -
06-27-2008 12:53 AM
Dear Sir,
there may not be a problem of Channel port or anything.
My 2960 Sw is connected 3/16 and 5/15 port.
I found one thing that is Fiber Cabel is MM on core side patch cord is MM and GBIC is SM.
On 2960 side Patch cord is SM and SFP Module is SM.
06-27-2008 01:05 AM
Well then it looks pretty much right. For MM fiber cable, you should use SX GBIC and for SM fiber cable, you should use LX GBIC.
What is the error disable message that you are getting from the logs ?
- sn -
06-27-2008 01:10 AM
Dear Sir,
On Core switch I had tried to see log by sh log or sh logging but I do not find any logs for it.
If the other way to see it pls tell me... Its a CAT OS
06-27-2008 01:12 AM
To see the logs on a CatOS, do a 'sh logging buffer' ....
I presume that the fiber connection is up on both the 6506 and the 2960 switch right ? You should see a green light on both ends of the link.
Cheers,
- sn -
06-27-2008 01:15 AM
Ya its working but after some time by continuos 4- 5 Time poing up and down will put port in errordisable mode.
Same senario is working when 3500 series with SM GBIC is there instead of 2960.
And ya I do not find any logs on Core for error Disable
06-27-2008 01:24 AM
If as you say the problem is due to the link going to error disable state, the logs will usually tell why it went to error disable state ... .ie) due to udld, bad ether channel configs etc ....
So until you determine the root cause of the issue on why the switch went to error disable state, this issue will keep on coming.
One way to overcome this is to disable error-disable state, but it is not advisable as error-disable is to protect the switch from becoming unusable.
cheers,
- sn -
06-27-2008 01:37 AM
Dear Sir,
After monitoring it is found that After flapping port goes in to Errordisable mode.
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: