PP 8600 Error

[leomavv]

Dear All

Can anyone suggest what does the following warning mean:


[05/30/03 15:26:43] Link Down 3/6.Port is an access port
[05/30/03 15:26:43] Link Down(3/6)
[05/30/03 15:26:43] WARNING Task=tMainTask Shutdown port 3/6 due to excessive control frames multicast 0, broadcast 56122 packet per second

How does the excessive control frames generated?

Regards

Mavvy

 

[mucka]

hi mavvy , yes we seen the same message on one of our 8600,s .We spoke to nortel and it seems we have had a broadcast storm probably due to a LOOP between two seperate lans .we had dumb 3com hubs and when we replaceed them with intelligent switches that is when we got the error.the passport uplink port sees the storm and then shuts down the port.
You will need to find & remove the loop then disable and re-enable the port OR re-boot the 8600 to bring the port back up.This was my understanding of the problem any way if anyone has any other ideas I would love to hear them .

thanks

mucka

 

[PanDa1]

Hallo maavy,

that´s right this message occures when an Broadcast or Multicast-storm ingress an port. The port shuts down and can protect the CPU of the passport. This is not a bug but a feature called cp-limit includet at sw-version 3.2.2
This feature is configureable per port basis or you can it disable per port basis´. The default settings are enabled.

thanks
panda1

 

[leomavv]

Thanks a lot for the useful info

Regards

Mavvy

 

[ecuamer]

I ran through that error message last month when some guys in my group were working on a BPS2K stack and some how when they brought that stack up it knock down the uplink port to the Passport . I was able to get back to the remote Passport by disabling and enabling the fiber port on the Adjacent passport . Nortel has set a rate-limiting feature to protect the cpu from excessive ingress broadcast but you can set your own parameter if you so desire.

 

[Ewey]

This can be a pain./
We have had some genuine server generated traffic hit this threshold -10Kpps- they were Ghosting a lab full of dumb client machines.
If your problem is caused by genuine traffic -and not miss-configuration loops, then you can soft set the threshold to 9Kpps.
This will stop excessive B'cast/M'cast frames, but you don't need to reset the port afterwards, as the hard-coded threshold is never hit.


Cheers,
Ewan