CSCur05027 - Loopback error occurs

[AndyFeng]

I experienced a major network outage last weekend：

The uplink ports of many access switches become err-disable. Each access switch is connected to the core switch through port-channel.
When I checked the logs, I found:
1. The faulty access switch has the following prompts:

ETHCNTR-3-LOOP_BACK_DETECTED & PM-4-ERR_DISABLE loopback error.

2. The core switch is connected to many access switches, but the switches with uplink port err-disable only have the following versions:

Switch Ports Model              SW Version            SW Image                
------ ----- -----              ----------            ----------              
*    1 52    WS-C2960X-48LPS-L  15.0(2a)EX5           C2960X-UNIVERSALK9-M    

Switch Ports Model              SW Version            SW Image                
------ ----- -----              ----------            ----------              
*    1 52    WS-C2960X-48FPS-L  15.0(2)EX5            C2960X-UNIVERSALK9-M 

The remaining normal switch system versions are: 15.2(2)E6 or 12.2(55)

Question: I don't know if this fault is related to bug:CSCur05027, because the description of bug:CSCur05027 does not mention 15.0(2a)EX5/15.0(2)EX5, only 15.0(02)EX01

Supplementary fault information:
The failure happened suddenly over the weekend, and no one modified the switch configuration.
The failed switches are spread across different physical locations

Hope someone can help me find the reason, I'm very confused. Thanks!

[balaji.bandi]

as per the message looks like some where there was STP Loop

what is your core switch model and code ?

the one error disabled and one not effected same L2 etherchannel to core switch ?

how is config looking like some new version has feature err disable recovery ?

Note : also check any keepalive config issue

[AndyFeng]

The model and version of the core switch are:
Cisco C6807-XL Version 15.3(04r)SYS

The port-channel and physical port of the core switch are configured with keepalive by default, and the port-channel of the access switch is also configured with keepalive by default, but the physical port of the access switch displays keepalive not set.

I'm not sure if it's related to CSCur05027, because CSCur05027 doesn't mention 15.0(2a)EX5. But this failure happened suddenly, and only 15.0(2a)EX5 and 15.0(2)EX5 failed, and other access switches connected to the core switch using port-channel were normal.

[balaji.bandi]

If possible upgrade the code....suggested or work around ? (may be effecting with bugs)

also i compare config working vs not working. make changes as per the environment and test it.

if this is 1 or 2 devices you can apply quick fix as suggested , if this is effecting many devices ? you need to look there the cause of the issue.