MGCP IP phone keeps recycling power

[telnettech]

Can someone possibly help me figure out what the heck is going on.

We have numerous phones that someone can be on a call and place it on hold or even hanging up and the phone will just reboot. I have checked cable, jacks, switch, swapped out the phone, moved to another switch and even pulled a new cable between desk and switch. It doesnt do it all the time and will sometimes not happen for weeks. Then all of a sudden, 2, 3 maybe even 4 times in 1 day.

These are the Cisco IP 7960 models and the firmware is POM3-07-5-00 using the MGCP protocol and getting power from a switch(POE).

thanks
Brian

To error is human.....if the machine doesnt work, then KICK IT !!!!!!!!!!!!!

 

[telnettech]

anybody?




To error is human.....if the machine doesnt work, then KICK IT !!!!!!!!!!!!!

 

[whykap]

Could someone been making changes to these phones and that's why they are resetting?
Is this a remote site?
Are they all on the same switch?
If so it could be a switch issue. It can also be an arp issue.
There are too many reasons a phone will reboot that I can only speculate without knowing your environment.

 

[telnettech]

whykap

they are on the same switch and it is not a remote site. There are no changes happening at the time of the reboot. We have checked the SNMP of the switch and there is nothing out of the ordinary traffic that is happening at the time of the reboots. We have checked cables, jacks, repunched down the outlet, even replaced the cabling and the phone to no avail. like i said, it is random and only happens when they are hanging up the call.

I will have to check the ARP messages for the phone we are monitoring.

Any other suggestions would be appreciated

Brian

To error is human.....if the machine doesnt work, then KICK IT !!!!!!!!!!!!!

 

[whykap]

Are all the phones on that switch behaving that way? Do you have other switches that are the same type and have the same IOS that work fine?
I am just guessing at this point but you cannot exclude the switch just because there are no snmp logs.

 

[telnettech]

whykap,

Yeah there are other switches of the same type and same IOS that the issue is not happening.

The phones that reboot are on the same switch cause they are in the same area. These are POE switches and we have tried to disable the POE and put power bricks out but that doesnt make a difference.

I have a Wireshark going on at 1 of the phones and am waiting for a call to say that it happened again so that I can look at the packets and see what is going on at that time of reboot.

Just was looking for some guidance on what I may not be thinking about that would cause this issue.

Brian

To error is human.....if the machine doesnt work, then KICK IT !!!!!!!!!!!!!

 

[telnettech]

ok i have a PCAP and see that the phone is using RSIP at the time of the restart....the edgemarc switch is responding to this as a protocol error........


any ideas?


Brian

To error is human.....if the machine doesnt work, then KICK IT !!!!!!!!!!!!!

 

[whykap]

Are the phones registered to call manager as SIP or SCCP?

 

[telnettech]

We are not using a CUCM.....we have them registered to s Sylantro softswitch.


Brian

To error is human.....if the machine doesnt work, then KICK IT !!!!!!!!!!!!!

 

[whykap]

So the edgemark gateway talks to the softswitch on the provider side and also the phones register to it using SIP protocol. Since the gateway is giving errors on the registration request for these few ip sets but not the rest of the phones I would thing it is a configuration setting on these particular cisco phones or it could also be a firmware mismatch.
you need to contact sylantro and see why the gateway is spitting out these errors.

Unfortunately I don't know anything about Sylantro as it not a CISCO product so I can't help you there.

 

[telnettech]

found this issue and believe have resolved it. The Edgemarc was not liking the RSIP restart method of keepalive (it didnt know what to do with it). So after failing 3 times, the phone was sending a notify and wasnt being answered by the Edgemarc. Cause of this failure, it would send another RSIP but the restart method would be disconnected and the Edgemarc liked that. So therefore the phone would reboot to get a new IP address and re-register.

I changed the keepalive in the TFTP file to be 'disabled' and now I dont get the keepalive RSIP and there is no restarting of the phones.

Thought I would post a solution to this issue for future use.

Brian



To error is human.....if the machine doesnt work, then KICK IT !!!!!!!!!!!!!

 

[whykap]

out of curiosity where is the TFTP file you mentioned?

 

[telnettech]

It is on a TFTP/DHCP server in our network. It is named as MGCMACADDRESS.cnf...... we have a default cnf file that is used if a more specific cnf is not available.

When the phone pulls DHCP, it is told to go to this ip address and get its configuration based on the mac address. If it doesnt find the MGCMACADDRESS.cnf, it uses the MGCdefault.cnf

Brian

To error is human.....if the machine doesnt work, then KICK IT !!!!!!!!!!!!!

 

[whykap]

Ok thanks. It is strange however that only some of the phones would behave that way and not all.

 

[telnettech]

When I looked at the pcap for a phone that it was happening to and a phone that it wasnt, the only difference was the RSIP message was no the phones that were rebooting. I can only assume that somehow the phones that were rebooting had the keepalive setting enabled in the software and since we had it commented out in the cnf file(wasnt enabled or disabled), that it was using what was already previously programmed though it should have went back to 'disabled' per the Cisco documentation in the software of the phone(default setting of the software). This is the only thing I could come up with to explain the issue.

Brian

To error is human.....if the machine doesnt work, then KICK IT !!!!!!!!!!!!!