[cisco-voip] 7900 series and ARP

[Pawlowski, Adam]

Absolutely. Like several other quirks and bugs, it may turn out that there is a very specific set of circumstances which contributes to this problem, which may likely disappear after a change like a new load, one piece or another rebooting, something else on the network leaving, etc.

Thanks though for the help and I will post if I find anything new.

Adam Pawlowski
CIT/OSS Repair Services
University at Buffalo

> -----Original Message-----
> From: Wes Sisk [mailto:wsisk at cisco.com]
> Sent: Thursday, July 01, 2010 3:07 PM
> To: Ryan Ratliff
> Cc: Pawlowski, Adam; cisco-voip at puck.nether.net
> Subject: Re: [cisco-voip] 7900 series and ARP
> 
> And after that it will likely be best to proceed through a formal TAC
> case.  There isn't a plethora of information available in this area so
> it will take some dedicated time and communication.
> 
> /Wes
> 
> On Thursday, July 01, 2010 2:58:27 PM, Ryan Ratliff
> <rratliff at cisco.com>
> wrote:
> > Before you do anything you need to upgrade the phone firmware to the
> latest load.
> >
> > -Ryan
> >
> > On Jul 1, 2010, at 2:06 PM, Pawlowski, Adam wrote:
> >
> > The phones are 7961G-GE's and 7941G-GE's. Load is 8.5.3S . I have a
> partial capture, not being mindful of my buffering when I was able to
> reproduce it, I missed call setup. Basically there is a bunch of
> broadcast ARP request from the destination phone to the originating
> phone but it does not reply. I toggled GARP from off to on, which
> shouldn't have really made a difference, and following is the start of
> the next call which then it answers the first request and the call
> begins.
> >
> > Except for that one instance I haven't been able to reliably
> reproduce it, only receive reports of the issue, and then I can see on
> the console log, from the phone, the ERR about not receiving any RTP
> packets. Of course within the last couple of loads the log just fills
> up with "CDP-D: holdThrd SendingCmd:proto:1 port:0" leaving us with
> only a short time window to observe, if the user happens to call in as
> soon as it happens.
> >
> > If there is a theoretical scenario it may be easier to try and
> reproduce it here in a test environment, barraging the phone with some
> variety of traffic, if need be.
> >
> > -----------
> >
> > Adam
> >
> >