Unit in switch stack gets sometimes down (2550T)

[coreysic]

Hi everybody,

I work for a French school with a LAN network composed of several Nortel 2550T in stack (bunch of 3 to 5 units).
Everything worked fine on the network until one day where some switches went down without doing anything.

Some 2550T stacked (one stack of 3 units and one stack of 4 units) went down and by examining logging message we figured out 2 critical informations :

• 1st the switch send that message : 02:26:10 GMT+01:00 7 NVR Sw Exception: Task tCDTImc, Type Data Access, PC 0x00958194, SP 0x03005ef0
• 2nd the switch sends that message : 02:27:24 GMT+01:00 6 NVR Trap:  s5CtrUnitDown

In fact on both stack the unit #1, which is the base, goes down and all the stack gets paralyzed. We need to reboot all the stack.

Both stack have :

• FW:1.0.0.14
• SW:v4.3.0.004

Has anyone faced that kind of weird issue ?
I don't understand the message Type Data Access,  PC 0x00958194, SP 0x03005ef0. Has anyone a clue about this message ?

Thanks.

Maxime

[Flintstone]

Hi Maxime and welcome to the forum,

I think you will need to raise a call with Avaya?

CheerZ and good luck

[coreysic]

In fact the school bought those switches using a revendor and not to avaya directly, any assistance or firmware upgrade is really expensive.

Do you know a way to contact avaya support without asking the revendor (We don't have any sold number required by avaya when registering on their website).

I know that without an avaya account with a sold number you can't access their technical assistance and firmware upgrade. Is that right ?

[Flintstone]

Hi Maxime,

I believe unless you have a maintenance contract you will not be able to get technical assistance or firmware.  If you call anyway, they might ask for your credit card details before continuing?

I would try your 'revendor' first to see what they can do for you as they did sell the switches to you and might already have the newest firmware availalble?

CheerZ

[Michael McNamara]

It appears (from the logs) that the issue is a software exception (Sw Exception). There's some problem that causes the switch throw an exception and reboot in order to recover. It looks that you are only 1 software release behind the current release of 4.4 and I don't see anything in the software release notes that leads me to believe that you would be any better off with an upgrade.

https://support.avaya.com/css/P8/documents/100145944

If this happened just once I think you can just ignore it... similar to why Windows needs to be rebooted every few days.

Cheers!

[coreysic]

It appears (from the logs) that the issue is a software exception (Sw Exception). There's some problem that causes the switch throw an exception and reboot in order to recover. It looks that you are only 1 software release behind the current release of 4.4 and I don't see anything in the software release notes that leads me to believe that you would be any better off with an upgrade.

If this happened just once I think you can just ignore it... similar to why Windows needs to be rebooted every few days.

Cheers!

In fact that kind of problem appears randomly about every 2 days.

We enable 802.1X on all access ports and it appears that bdpu-filtering functionality doesn't work anymore if you enable 802.1X (we made a bunch of tests on a fresh 2550T). Maybe a bridging loop appears on one port causing a SW Exception.

[Flintstone]

Hi Maxime,

As Michael mentions it does sound more like a software issue than a hardware issue? 

Try removing your bpdu-filtering functionality, especially as it doesn't work anymore to see if that improves things?  With 802.1X enabled you probably don't need bpdu-filtering anyway?

CheerZ