01-20-2014 10:05 AM - edited 03-16-2019 09:20 PM
Hello,
I have an CUCM 9.1 cluster. I installed a 2901 router as an MGCP gateway in SRST mode at a remote site. Everything seems to be working ok, but I can't use the Call forward all option on the phone. When I try to forward my ext to my cell phone, the call initially goes through. But after hanging up, both of my fxo ports stay off hook and have to be reset to allow for dial tone again. I can forward calls internally to other internal exts with no problem. I have tried changing several voice port options, like battery reversal, timeouts call disconnect and timeouts call wait release. What could be causing this issue?
voice-port 0/1/0
supervisory disconnect dualtone mid-call
timeouts call-disconnect 5
timeouts wait-release 5
connection plar opx 5000
description xxx-xxx-xxxx
caller-id enable
!
voice-port 0/1/1
supervisory disconnect dualtone mid-call
timeouts call-disconnect 5
timeouts wait-release 5
connection plar opx 5000
description xxx-xxx-xxxx
caller-id enable
!
!
Thanks
Bill
01-20-2014 11:10 AM
Is this behavior occurring only for call forward all or normal calls as well?
You mention SRST, how is SRST related in this issue?
Chris
01-20-2014 11:44 AM
I don't think SRST is related to the issues. I just mentioned that it was in use. It only happens on Call-Forward all.
01-20-2014 12:25 PM
It looks like router is not able to detect the disconnect tone when call is getting forwarded. Disconnect tones are detected by router DSP. Generally router will not allocate DSP for POTS to POTS (hairpin) call leg (in your case between FXO port 0/1/0 and 0/1/1) so because of it FXO port remains OFF HOOK. To confirm dsp is getting allocated or not run "show voice call status" and check DSP channel is allocated or not.
If it is none then to add DSP to POTS to POTS call you should configure below command
voice-card 0
no local-bypass
Regards,
Mohit Singh
01-20-2014 12:49 PM
Hello and thanks for your reply.
It appears dsp resources are allocated
CallID CID ccVdb Port Slot/DSP:Ch Called # Codec MLPP Dial-peers
0xAFE 3B 0x22DB17A0 0/1/1 0/1:1 5000 g711ulaw 2/1000
0xB00 11F2 0x3D387D34 0/1/0 0/1:2 * g711ulaw 0/0
2 active calls found
I tried adding the commands anyway but I stillget the same results.
voice-card 0
no local-bypass
Thanks
Bill
01-20-2014 01:28 PM
Can you provide "debug vpm signal" and "debug mgcp packet" for a test call? Make sure you collect both debugs together.
Also I want to know if forwarding is removed and IP phone answers the call then in this scenario FXO port goes to ON hook state as soon as PSTN caller disconnects the call?
Regards,
Mohit Singh
01-22-2014 06:46 AM
Hello,
Thanks for your response. Sorry for the delay. I haven't been on site. I will get the debugs on Friday,
Thanks
Bill
01-22-2014 09:41 AM
Guys,
The POTS to POTS hairpin will not apply here because at the moment the GW uses MGCP, the call will still have a VoIP leg so no POTS to POTS hairpin will happen.
Now, since the call has both ends connected to Telco that means the Telco will be the one initiating the disconnect which will need to be detected by the DSP on the FXO and then trigger the disconnection to the other FXO.
With the debugs Mohit requested we should be able to see if the Telco sends at any point a disconnect supervisory (battery reversal or power denial).
If none of this disconnections are detected, then the DSP would only be able to disconnect the call by listening to disconnect tones so a PCM capture would need to be done in order to check of any disconnect tone is coming from Telco.
Hope it helps.
~ Luis Ramirez,
01-24-2014 06:44 AM
01-24-2014 07:09 AM
Hi Bill,
I checked the logs and I don't see FXO is able to detect supervisory disconnect from PSTN. Also you mentioned that in normal call also port is not going to "On hook" state when call is disconnected. Is "battery-reversal" disabled on your FXO port? if yes then enable it. If still your port does not go to on hook state then collect PCM capture for a normal call. I have attached document which explains how to collect PCM capture.Steps are specific to IOS version so run the commands supported on your IOS version. Also share the FXO port configuration.
Regards,
Mohit Singh
01-24-2014 07:11 AM
Bill,
The call connected on the second port right here.
011812: *Jan 24 09:03:41.119: htsp_process_event: [0/1/0, FXOLS_WAIT_CUT_THRU, E_HTSP_VOICE_CUT_THROUGH]fxols_handle_cut_thru
And there is nothing being detected from Telco (power denial or baterry reversal).
The call goes down until the port goes down (due to a forced shut down done by CLI).
011841: *Jan 24 09:04:17.947: htsp_process_event: [0/1/0, S_DOWN, E_HTSP_RELEASE_REQ]act_disc_conf
011859: *Jan 24 09:04:18.923: %LINK-3-UPDOWN: Interface Foreign Exchange Office 0/1/0, changed state to Administrative Shutdown
It was connected for around 39 seconds and nothing was detected by the DSP.
With that information we know Telco is not sending the power denial or baterry reversal so you would need to capture a PCM in order to find out of the Telco is sending a disconnection tone.
01-24-2014 07:18 AM
Thanks for your responses. Battery reversal is disabled. I can re-enable it and test.I will do a packet capture and post.
I appreciate the help
Bill
01-24-2014 07:57 AM
I resolved my issue. I called my telco provider and they offered an automatic disconnect feature. I had them turn it on and everything is working now. Thanks for all the input
BIll
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide