[cisco-voip] 7940/60 dropping network connectivity

[Wes Sisk]

Jim, do your phones just suddenly unregister or is it only when they  
are being administered?  When they re-register to CM what do they  
report as their last unregister code ( cmd prompt in the ccm trace  
directory c:\prog*\cisco\trace\ccm, 'findstr Last= ccm*.txt > last.txt')

2 points come to mind:
1. all recent 7940/60 loads are programmed to keep the internal  
switch online unless there is power loss or hard reset.  Even if the  
phone software crashes we preserve the switch so we can preserve  
inline power and therefore capture a stack trace on the phone that  
you can see on phone's web page.

2. there have been several recent changes in the switches to fix  
speed/duplex negotiation and modify buffer sizes on the phone because  
of excessive dropped frames on the PC port.  One issue is  
characterized by the network interface bouncing up/down between the  
phone and PC.  The other is slow throughput to the phone.

What phone load are you running?  Do you see network connectivity  
bouncing, no throughput, or slow throughput?

/Wes

On Apr 20, 2006, at 4:44 PM, Robert Singleton wrote:

On Thu, 2006-04-20 at 15:18 -0500, Carter, Bill wrote:
> The primary cause of this problem is the phones 3 port Ethernet switch
> was software based on these models.  The 7912G-A converted to a  
> hardware
> based switch like the 7940G and 7960G.

In a related issue, I have some 79[46]0 phones that seem to momentarily
lose the user access port very briefly under some reset conditions.
Sometimes, it's as simple as making a small change to the phone that
requires a soft reset. More critical issues occur when the phone loses
contact with CallManager and keeps trying to re-home.

We are a wholesale supplier and our typical branch office is designed,
quite intentionally, to not be brought completely down by a loss of the
T1 to corporate. All sales and inventory functions take place on the
local printer. If we lose connectivity to corporate, we lose telephones,
but business marches on unaffected.

Well, except that whenever an IP phone is attempting to rehome to the
CallManager (which is at corporate), it will reboot. During this reboot,
there is enough interruption of the swiched traffic through the phone to
disrupt the telnet or ssh connection from the user's PC to the local
server.

Worse, not all phones do this and even the ones that do, don't seem to
do it every time they lose connectivity to CallManager.

Any ideas?

Robert

_______________________________________________
cisco-voip mailing list
cisco-voip at puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-voip