2960s stack non-stop reloading

[Anuar Shahrin]

Hi All,

Anybody experience this issue before?

Switch model: WS-C2960S-48FPD-L

Mix stack with 2960-X

We upgrade the IOS software from C2960S-UNIVERSALK9-M), Version 15.2(2)E6 to version C2960S-UNIVERSALK9-M), Version 15.2(2)E7

After reboot, the switch keep on rebooting with a traceback message.

 21:06:19 MAL Fri Mar 9 2018: Unexpected exception to CPUvector 2000, PC = 235615C
-Traceback= 0x235615Cz 0x16A92A8z 0x16DD360z 0x169D90Cz 0x21A936Cz 0x21A56ECz

we tried re-seat the stack module, unplug and plug the Stack cable, non successful.

Finally, we take out the stack module and reboot back to the previous IOS version.

Anybody faced this issue before? I saw other discussion regarding the stack issue, most of it is on older version of IOS image. I would imagine the issue should be solved in the newer IOS version.

Switch details:

Motherboard assembly number     : 73-12561-09
Power supply part number        : 341-0382-02
Motherboard serial number       : FOC16515W5H
Power supply serial number      : DCA1638V3CV
Model revision number           : E0
Motherboard revision number     : A0
Model number                    : WS-C2960S-48FPD-L
Daughterboard assembly number   : 73-11933-04
Daughterboard serial number     : FOCXXXXXX
System serial number            : FOCXXXXXX
Top Assembly Part Number        : 800-32643-03
Top Assembly Revision Number    : A0
Version ID                      : V03
CLEI Code Number                : COMGH00ARC
Daughterboard revision number   : A0
Hardware Board Revision Number  : 0x01

[Reza Sharifi]

Hi,

Traceback messages are usually an indication of a software crash.  If you can't stop the reload by pressing the mode button, I recommend you open a ticket with TAC and have them look at it.

https://www.cisco.com/c/en/us/support/docs/switches/catalyst-2950-series-switches/41845-192.html

HTH

[Anuar Shahrin]

Manage to stop the reload using mode button. Now all of my stack module is removed and we revert back to the old IOS .E6

[bartosikr1]

Hi Anuar

Did you ever find resolution to this issue? We attempted to upgrade out CISCO 2960s stack to 152-2.E7 and experienced the same non-stop reloading. We had to revert back to the original 15.0(2a)SE9 to get the stack working again.

Do you have any recommendations?

Thanks,

Richard

[Leo Laohoo]

Create a new thread.
Post the boot-up process so we can determine what is going wrong.

[Anuar Shahrin]

Hi Richard,

Unfortunately, the only solution was we revert back to E6. Does your stack showing any traceback after the boot?

No issue with our 2960-X running on E7.

[Leo Laohoo]

I'm with Reza. I'm using E7 and my stack is fine. Raise a TAC Case and get them to find the root cause of the crash.

[FTBZ]

Same issue here with 15.2.2 E8, reverting to E6 solved the issue.

[Anuar Shahrin]

Yeah, the ony solution was revert back to E6

[FTBZ · Accepted solution]

We opened a TAC case, seems linked to the bug CSCvf46629. This is a VTP issue, the switch must not be configured in VTP Mode Client at boot. Changing the switch to VTP Mode Transparent at boot and set VTP Mode Client after the boot will solve the issue.

The case is not accurate on which version is solving the problem. Someone can test? I've no more similar devices to try.

@Leo Laohoo can you confirm that you're using Transparent?

[Anuar Shahrin]

We are running on client mode for all our access switch. Let me give it a try during our MW.

[bartosikr1]

Thanks! I would be very interested in your findings!

Have a good day!

[Anuar Shahrin]

I've change the VTP mode to transparent and upgrade to E8 without issue on a stack.
Already upgrade 2-3 stack in my network and it works.

Thank you so much all.

[FTBZ]

For information, the client mode isn't working after a reboot. You need to keep the Transparent mode if you need to reboot or upgrade again.

[FTBZ]

We tested the latest release E9 and the problem is definitively solved. Moving from E6 to E9 is working perfectly keeping the VTP in client mode.