Phones dropping out

[Rich14]

3300 CX 11 Release 4.2 SP2 PR1
5340e Phones
Dhcp supplied by customer server. (Flat network, No Vlans)

All phone at random intervals drop out and show "contacting server" and then re-boot back into service after approx 20 mins. They can drop out mid call. If the phone is unplugged and re-booted it comes straight back into service. Some times a phone can drop out 3 times in a morning and then work o.k. until the next day. The following log is generated each time it happens.

Heartbeat Server
ICP has lost contact with (192.168.1.199 (08-00-0F-77-BB-99)), network driver cluster pool free: 3004 and low water mark: 2883
Main
HeartbeatRecvTimeoutHandler.cpp;279.

Any ideas to the cause of this would be appreciated.

 

[MitelInMyBlood]

This can be tough to find. However, my bet is that it's a network problem and not a phone system problem. With VOIP your phone system controller is only present for the call setup. Once a VOIP call is established between two endpoints, you could literally turn off the main control and the call would stay up.

We once fought a similar problem and to our shock discovered it was due to our reconfiguration of the ARP timers to re-ARP the network at 5 minute (600-second) intervals to accommodate the server room's lust for swapping out equipment and always wanting to re-use the same IP address. Once we set the ARP timers back to their system default value, the problem disappeared.

Not suggesting that this is your problem, just merely wanting you to look more closely at your network than at your phone system. If you have any network design issues you will begin finding them once you put in a VOIP phone system. Everyone will blame the phones when the problem is the network. (They blame the phones because that us what's being affected even though it's the network causing it)

Original MUG/NAMU Charter Member

 

[danramirez]

I agree with MitelInmyblod, i would say that is a network issue.

If i'm not wrong there is a tool called the network analyser, have you use it? I think you would have to program your laptop's IP address within option 125 so your laptop can take statistics from the phone under study.

Maybe the network analyser can provide additional information that can point you to solve your network problem.

Regards,

Dan

http://www.hatkana.com

http://www.wtelecom.com

 

[danramirez]

The tool is called IP Phone Analyzer and can be downloaded from MOL

http://www.hatkana.com

http://www.wtelecom.com

 

[LoopyLou]

Other then a software bug it sound like the heartbeat signal is either not getting to the 3300 or the controllers response is not getting back to the phone in a timeley manner.

I'd tell you a UDP joke but I'm afraid you won't get it. TCP jokes are the best because you always get them.

 

[Rich14]

Thanks for suggestions.
I tried using the Phone Analyzer but couldn't get the phones to register with it. (put 131 in customer dhcp server and re-booted phones).
Then tried wireshark, (struggling to use it). I may have got a tie up with ICMP, seeing ICMP messages and ocassionally a phone will lose its heartbeat at about the same time, but not every time.
Would ICMP broadcast be stopping heartbeat, why different phones each time ?

 

[bobcheese]

Loads of issues with ICMP redirect messages causing reboots in the past:

http://mitelforums.com/forum/index.php?topic=1692.0

http://mitelforums.com/forum/index.php?topic=836.0

http://www.tek-tips.com/viewthread.cfm?qid=1605849

Worth turning redirect off to test