cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1652
Views
0
Helpful
6
Replies

Packet loss to the 3850 access switch

Hello Everyone,

 

We could see  there is a packet loss to the access switch. I will say the setup below. Please help me to fix this issue.

 

Core switch(6504) [Gi1/3/8 & Gi2/3/8]<--> [Gi1/1/1 & Gi1/1/2] 3850 Access Switch[Gi1/0/48] <--> [Gi0/47]Downlink Switch

 

This is the setup. Note connectivity between Core switch and access switch is Fiber link. and there is no Port-channel configured. Then i could see Loopguard enabled globally in access switch and getting Loopguard blocking and unblocking alerts continuously. 

 

I have tried shut one of the uplink in Core switch. Like Gi1/3/8 and traffic passed via Gi2/3/8 still faced same issue. Packet loss. Vice versa same issue. Then verified the Transceiver details in interfaces no alarm or issue or error on fiber link. Please help me on this. 

 

Regards,

Chandhuru

Thanks and regards, Chandhuru.M
6 Replies 6

balaji.bandi
Hall of Fame
Hall of Fame

Where is the packet loss you see, end device side ? or gateway ?

 

Can you post the relavant config of Cat 6504 and Cat 3850  (if the no port-channel configured) how is these 2 Links connected using Trunk ?

 

Then that where i see the issue, is there any reason for that as trunk ? since there is Loop so you see block and unblock ports due to some convergence in network (this is my assumptions)

 

Also can you post the logs you see that block and unblock Logs  here ?

 

show version

show run interface  gi 1/3/8

show run interface  gi 2/3/8

show  interface  gi 1/3/8

show nterface  gi 2/3/8

 

same information from Cat 3850

BB

***** Rate All Helpful Responses *****

How to Ask The Cisco Community for Help

Thanks for the update Balaji!!!

 

Where is the packet loss you see, end device side ?  or gateway ? Access Switch and downlink Switch

 

Can you post the relavant config of Cat 6504 and Cat 3850  (if the no port-channel configured) how is these 2 Links connected using Trunk ? Attached log file

 

Then that where i see the issue, is there any reason for that as trunk ? since there is Loop so you see block and unblock ports due to some convergence in network (this is my assumptions) Yes you are right but if i shut one uplink in core switch also seeing same drops and same issue

 

Also can you post the logs you see that block and unblock Logs  here ? Attached in word file

Thanks and regards, Chandhuru.M

Looks you have too much QOS Configuration configured on the Interface. and also i see some Drops on the interface not major.

 

For testing - Take one of the interface default the interface config, make it simple config and test it is good idea as below :

 

 

On Core switch remove cable 2/3/8

config t

!

default int gi2/3/8

interface GigabitEthernet2/3/8
description Uplink to Access Switch
switchport
switchport mode trunk

no shutdown

!

on the access side :

 

config t

!

default interface GigabitEthernet1/1/2

!

interface GigabitEthernet1/1/2
description *Uplink to Core Swt 2*
switchport mode trunk

no shutdown
end

 

 

Now connect the cable between these devices, remove connection from  Core Gi 1/3/8

 

what is the out come ?  (based on the results we fine tune the requirements.

 

 

 Attached in word file

this is missing here can you update the post, let me read the logs

BB

***** Rate All Helpful Responses *****

How to Ask The Cisco Community for Help

Thanks Balaji!!!

 

Attached here as well log file

 

Can we expect any Bug on this? Because switch is currently running on 3.2.2.

 

I will try the suggestion as you said. I dont think it is fix. Because same setup for 2 more switches those are not seeing any latency. Same interface config with same Qos. Anyway let me try as you said. and keep you posted.

Thanks and regards, Chandhuru.M

Hi Friend,
still face issue here ?
if yes then can you check downlink SW elect as root for this domain, Access SW see new root and this make port "loop guard" error.
please make Core SW priority prefer than that of downlink SW and see result.

Sorry for the delay response!!! Thanks for the reply MHM!!!

 

Issue has been resolved now!!! 

 

Looks like switch was hit with BUG. Because we tried rebooting the switch has fixed the issue. Now we are not seeing the LOOPGUARD error on logs. 

 

Note: Not sure the BUG details. We suspecting unknown BUG. If anyone aware of this BUG details. Please share. 

 

Switch Model: Cisco 3850

Version: 3.2.2

 

Regards,

Chandhuru

Thanks and regards, Chandhuru.M
Review Cisco Networking products for a $25 gift card