Authentication Failures

[TerryG]

I have encountered an annoying "problem" with my new 3CX 12 SP6 installation. Over the Christmas holiday we replaced the company's old PBX with this and 40 Cisco 7941 handsets.
All the phones are registered and working however when looking in the server activity log It is full with repeats of this (for each extension). The errors seem to repeat every few seconds for each handset. I currently have 1213 pages of logs covering about 5 hours.

13-Jan-2015 13:47:55.302 [CM302002]: Authentication failed due to unidentified source of: SipReq: PUBLISH [email protected] tid=2d31ad45 cseq=PUBLISH [email protected]:5060 / 217097 from(wire)
13-Jan-2015 13:47:54.034 [CM102001]: Authentication failed for AuthFail Recv Req PUBLISH from 192.168.40.19:53144 tid=6c86fe4c [email protected]:
PUBLISH sip:[email protected] SIP/2.0
Via: SIP/2.0/UDP 192.168.40.19:5060;branch=z9hG4bK6c86fe4c
Max-Forwards: 70
Contact: <sip:[email protected]:5060>
To: <sip:[email protected]>
From: <sip:[email protected]>;tag=0024c4bf84c3ccdde9145066-be07ce5f
Call-ID: [email protected]
CSeq: 50969 PUBLISH
Expires: 2147483647
Content-Disposition: session;handling=required
Content-Type: application/pidf+xml
Proxy-Authorization: Digest username="202",realm="3CXPhoneSystem",uri="sip:[email protected]",response="8af256508896f5d5639bcd177f4cb360",nonce="414d535c0ac5b30a82:b517c0f4a8b004c912d4daea1b88f525",algorithm=MD5
User-Agent: Cisco-CP7941G/9.4.2
Event: presence
Content-Length: 465

10.0.0.210 being the IP of the 3CX server and 192.168.40.x being the vlan containing the ip phones.

We also have a number of Yealink W52P DECT phones which don't appear to be logging any errors.

So far I have tried different firmwares (8.5, 9.2 and 9.4) on the handsets, re-provisioning with default settings, downloading the latest provisioning template, checking the router to ensure all packets are being passed through correctly (as far as i can tell they are), adding a new extension on a new handset and have now run out of ideas.

The reason for looking at the log is some of the sales staff have commented that a small number of inbound calls have sounded "corrupted" for a moment during the call then resumed fine.

Is this something expected with the cisco handsets? A morning of googling only finds one reference to this error but with a 7961 handset and no solution.

Thank you in advance for any help you can provide.

 

[lneblett]

I don't know that this will help, but here goes:

I do not know how this phone was originally being used and if you had to reflash the firmware and if so, if that also removes ALL prior phone attributes and resets to a factory default.

If uncertain or know that it did not, then I suggest factory resetting a single phone as a test and then provisioning to see what happens.

For some reason it looks to me as though a MAC address is being used where the extension number or identifying number would normally go:

001ebe8f79ab@192.168.40.11:5060

I do not know that this is a MAC, but it is 6 bytes long and appears all Hex.

As a result, the system does not recognize the device and is refusing it.

Compare it with that of the Yealink.

I assume that you followed the blog instructions on how to make the 7941/61 compatible with 3CX. If not please review.

 

[TerryG]

Thanks for replying. You are correct that the part before the "@" is the mac address of the phone.
Using the handset on my desk as an example and cutting out the address lines of the error:

PUBLISH sip:[email protected] SIP/2.0
Contact: <sip:[email protected]:5060>
To: <sip:[email protected]>
From: <sip:[email protected]

The MAC address is my handset and the IP is either the phone or server (40.37 is the handset, 0.210 is the server)

As the to and from addresses are the same, is this the 3CX server trying to talk to itself?!?!

The handsets were all purchased as factory refurbished, unfortunately I am not aware of how they were previously configured. To see if it was causing the problem I have factory reset a couple of them but it hasn't changed the error. I have also tried both an unmodified and a simplified provisioning file on 8.5 and 9.4 firmware.
Just to satisfy myself that the vlan / routing isn't an issue i have connected a phone to the same network as the server, factory reset it and the error remains.

I have set them up using the instructions here: http://www.3cx.com/sip-phones/cisco-7940g-7960g/
also adding DHCP option 150 (the same as 66) after reading some replies to the blog.
I have found several working provisioning profiles for 3cx courtesy of Google including this one which i what i currently have loaded.
http://www.3cx.com/forums/using-a-cisco-7941-7961-phone-with-3cx-2561.html

I have been looking in the logfile viewer and can't find anything relating to the yealink phones (possibly as there are so many errors for the ciscos) to compare with.

Just in case this helps someone else, if you want to run a 9.x firmware on a cisco 7941/61 you have to set the proxy to "<proxy>USECALLMANAGER</proxy>" under siplines.

 

[lneblett]

Wish I could help you more, but I do not use these phones.

I did note that the 3CX instructions call for the supported firmware which is downloadable thru the 3CX blog site. In a comparison of the supported firmware between that of Cisco and 3CX it looks like 8.5(4) is the one, however, if you haven't already, I would use the 3CX download and try using only those files following a factory reset. You might also try to eliminate the VLAN so as to minimize any complications (try a demo system if need be). I noticed that you indicated you had tried 8.5, but wasn't sure of the context in which the attempt was made given that you also tried later versions.

 

[TerryG]

That's ok, any suggestions are welcomed. Googling problems has always provided a solution previously if i hadn't experienced it but this one has me stumped.
When I first noticed the problem I had the 3cx download 8.5.4 firmware installed. I tried the later versions to attempt a fix and as it doesn't seem to make any difference I have left them all on 9.4.2.
While testing today i have tried the 3cx firmware (8.5.4) and the latest offering from Cisco (9.4.2) with various different provisioning profiles and connected directly to the LAN.
I have found mention of this being a bug in 3cx version 7 which was fixed but no details of the fix.

It is fairly frustrating but it keeps me away from writing reports so i can't complain too much.
Cheers.

 

[lneblett]

Like your style - reports = bad, phone problem = good.

 

[TerryG]

More like: Reports - most boring job ever, phone problem - frustrating but not boring.

Anyway, progress of sorts. if i install a trial 3cx on a new VM and point one of the phones that causes issues with the production system at it, there are no errors logged. plug it back in to the production system and the errors return.
I am not sure what the difference is as i used the same source but tomorrows job is to restore a backup of the production system on to the test machine, point it at our SIP provider and try it with some test calls.

 

[TerryG]

After re-installing the phone system and restoring a backup, all the errors have gone away.
I'm not sure if that counts as a fix or not as I have no idea what the problem was but it's gone now so that's good enough for me.

Now to think about installing 12.5..................