[cisco-voip] 'Bad ARP Add' on 7900 series logs

[Pawlowski, Adam]

Brian,

I appreciate your insight

This is regarding the phone's software for voice functionality, and not applicable to traffic on the "PC" VLAN, right? 

Is there a sizing recommendation to the number of devices that are present on a subnet where it's phone-exclusive? The voice subnets are rather large. Perhaps, and I have to check on its current state, but disabling GARP would be helpful here? 

With a failure of adding an entry to the ARP cache, there could be a failure in operations with the phone with a call over some duration then? Are there other symptoms we could attribute to this?

Adam P
SUNYAB


> -----Original Message-----
> From: Brian Meade (brmeade) [mailto:brmeade at cisco.com]
> Sent: Friday, October 25, 2013 11:08 AM
> To: Pawlowski, Adam; cisco-voip at puck.nether.net
> Subject: RE: 'Bad ARP Add' on 7900 series logs
> 
> Adam,
> 
> When the phone creates a new socket for RTP TX/RX, it adds a permanent
> entry to the ARP cache to protect from the ARP entry timing out during the
> call.
> 
> After the call ends, this entry is supposed to be removed from the ARP
> cache.
> 
> That same "bad ARP Add" is printed if the port fails to be added to the
> permanent ARP cache or fails to be removed.
> 
> In your case, this is right after opening a new ingress channel via
> OpenReceiveChannel at the beginning of the call.
> 
> I would guess something is most likely filling up your ARP cache on the
> phones or something similar causing this to fail.  You should probably see if
> there's a any sorts of broadcast storms on your network that may be causing
> this issue.
> 
> Thanks,
> Brian
> 
> -----Original Message-----
> From: cisco-voip [mailto:cisco-voip-bounces at puck.nether.net] On Behalf Of
> Pawlowski, Adam
> Sent: Friday, October 25, 2013 10:23 AM
> To: cisco-voip at puck.nether.net
> Subject: [cisco-voip] 'Bad ARP Add' on 7900 series logs
> 
> Every so often I find myself trolling through the logs on our phones to ID or
> look for problems (which are rarely determinable here). I've done pretty well
> figuring out most of what I'm looking at to be able to deal with some of the
> basic stuff we'd run into , but, one message keeps catching my eye and I'm
> not sure what it is .
> 
> When looking at a call in the log, often, but not always, I'll see:
> 
> 2320: NOT 14:29:39.746772 DSP: STREAM- OpenIngressChan- ChanType 1,
> Remote (host 80xxxx85, port 585c), medType 4, Period 20, VAD 0, TOS b8,
> stream (113182798, 113182798) --> chan 0
> 2321: NOT 14:29:39.747517 DSP: STREAM- OpenIngressChan- mix (1, 3),
> dtmfpayloadtype 0
> 2322: WRN 14:29:39.748393 DSP: bad ARP Add, -126
> 2323: NOT 14:29:39.748959 DSP: Subtracted for CODEC[0] G.711 direction:1
> cost:13 old budget:87
> 
> Having, in the past, dealt with claims of dead air and no audio between sets
> (though not readily reproducible), I've seen this but can't figure out what it is
> supposed to be indicating, or if it's pointing to something that could be a
> problem in our configuration.
> 
> Anyone know what it's warning about?
> 
> Adam P
> SUNYAB
> 
> _______________________________________________
> cisco-voip mailing list
> cisco-voip at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-voip