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

[Pawlowski, Adam]

 Just so I didn't leave this empty for the list for future searching, I've noted
 when this has appeared in the console log, and pulled arp via "show arp" in
 the debug shell, after connecting to the phone via SSH.
 
 I only see the default gateway in the ARP listing there, so I don't know,
 perhaps this is just a cosmetic error. Doesn't seem to be causing me any
 trouble right now. This is on 9.3.1ES7S, so it's already in murky territory if I
 understand the ES designation correctly.
 
 Thanks again for your reply, Brian.
 
 Adam
 
 
 
> > -----Original Message-----
> > From: Brian Meade (brmeade) [mailto:brmeade at cisco.com]
> > Sent: Friday, October 25, 2013 12:04 PM
> > To: Pawlowski, Adam; cisco-voip at puck.nether.net
> > Subject: RE: 'Bad ARP Add' on 7900 series logs
> >
> > Adam,
> >
> > The ARP cache on the phone shouldn't be storing any entries from the
> > data VLAN I believe.
> >
> > I'm not aware what the limit is but I would say that I don't think the
> limit
> > would be hit from legitimate traffic.  My guess would be this is due
> > to
> some
> > sort of broadcast storm or a bug.
> >
> > The failure would most likely result in one-way/no-way audio for the call.
> >
> > If you log into the SSH shell on the phone, you can run "sh arp" to
> > take a
> look
> > at the ARP cache and see how large it is.  I'd suggest taking some
> > packet captures of the Voice VLAN to see what kinds of broadcast
> > traffic you're seeing.
> >
> > Brian
> >
> > -----Original Message-----
> > From: Pawlowski, Adam [mailto:ajp26 at buffalo.edu]
> > Sent: Friday, October 25, 2013 11:58 AM
> > To: Brian Meade (brmeade); Pawlowski, Adam; cisco-
> voip at puck.nether.net
> > Subject: RE: 'Bad ARP Add' on 7900 series logs
> >
> > 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
> >
>