Softphone & Iptouch reset issues S.O.S!!

juanattack · Post by **juanattack** » 20 Mar 2007 18:33

I'm having reset issues with softphones and iptouch phones since an OXE R7.0-f4401-19.-b was installed.

Here's a summary of the case:

- The IPtouch phones are using the PC port and it's ethernet port is connected to a CATALYST 3750 switch (L3/L2)

- Every phone gets its IP address from an external DHCP server.

- The Call Server, GD and GA are connected to the same switch through the uplink of the CS

- No VLANs or 802.1p/q is implemented, however doing a broadcast analysis and AVISO, the results came allright (no broadcast issues, or collisions, errors, CRCs, etc)

- All IP phones and boards are on domain 0, and using The IP QoS category 0 with the following parameters configured:
* UDP Lost: 30
* UDP Lost Reinit: 30
* UDP Keep alive: 60
I had configured the default values intially (7,7,15), but with these times we had less resets.

- Everytime they reinitialize the following incident appears on the OXE: 0378: IP phone signaling incident 6 on hardphones and 11 on softphones.

- Looking at the defence command on the Iptouch, the reset cause is UAUDP_LOST_CONNECT

- The last thing I did was making traces with ethereal on one of the IPtouch with the problem, and these were the results, just before the reset:
* The IPtouch sends UDP packets to the CS but never gets response (those should be the Keepalives encapsulated in UDP)
* I can see ARP requests from the CS to the phones, from the CS to other phones, from the GD to the phone, and from the CS to the GD ????????????
I don't know quite well how the MAC address table of the CS, GD and GA works, but it seems that somehow some entries get erased (FIFO maybe).

I've tried to install the libraries UAUDP.dll, to get a better look at the UA protocol, but it doesn't work on wireshark or ethereal's last version 0.99.0.

I would appreciate very much your help and/or comments.

P.s. Any chance someone has the ethereal V0.10.3 installer???

cavagnaro · Post by **cavagnaro** » 20 Mar 2007 18:50

1. Have you tried to connect directly to the OXE with an static Ip address?
2. How is the DHCP Server configured?
3. What data are you filling on TFTP Server?

zanex · Post by **zanex** » 20 Mar 2007 19:43

ive put TC798 on pastebin tc798.zip as it describes issues with upgrade to R7.0
Are the restarting constantly and never coming into service or just intermittently?
how many phones have you got ?
are the phones in the same subnet as the pabx?

juanattack · Post by **juanattack** » 20 Mar 2007 23:22

Cavagnaro:
1. No I havenÂ´t tried to connect directly to the OXE with a static IP address. However, CS, GD and GA, they all have static IP addresses.
2. ThatÂ´s a good point, I havenÂ´t checked if the DHCP lease expires...
3. In the TFTP address I programmed the CS IP address with the lanpbxbuild command.

Zanex:
- After the IPtouch resets it comes into service, and continues itÂ´s dialog with the CS and GD with no problem.
- IÂ´ve got around 40 hardphones and 60 softphones, and they are in the same subnet as the PABX.
- Thanks for the document, IÂ´ll check out if its original version was the releases mentioned, but as far as I know, this is a brand new OXE. Although it might have come migrated directly from France, you never know...

zanex · Post by **zanex** » 20 Mar 2007 23:55

When in service how is the voice quality
How often do they reset? Is it trafic related? Are the PC's on the same vlan? if so try vlan'ing a phone off and see if it resets

cavagnaro · Post by **cavagnaro** » 20 Mar 2007 23:55

juanattack wrote: 1. No I havenÂ´t tried to connect directly to the OXE with a static IP address. However, CS, GD and GA, they all have static IP addresses.

Try this and let us know how it behaves.

juanattack wrote: Although it might have come migrated directly from France, you never know...

? You have not uploaded the OS? It's always recommended that you upload latest patch and versions to be sure. We always had problems using pre-loaded versions.