Call does not disconnect when the pstn user hangs up

All replies

• 

  

  0

  Sign in to vote

  I have the same issue using an AudioCodes media gateway. It would really be great if there was a way to have it hang up automatically when the user on the remote end hangs up. If there is some way to make this work I would sure like to know.

  Friday, December 7, 2007 5:00 PM
• 

  

  0

  Sign in to vote

  I have logged an MS call as digging around a bit more the error code 

  <----SIP/2.0 481 Call Leg/Transaction Does Not Exist

  Call-ID:30474e34-8d50-463c-bd99-be67c20ec320

   

  Would seem to indicate that OCS is loosing track of the call which I can see was setup correctly.

   

  CSEQ: 591 INVITE

  CALL-ID: 30474e34-8d50-463c-bd99-be67c20ec320

  (I just hope the extra space isn't causing confusion)

  Friday, December 7, 2007 5:59 PM
• 

  

  0

  Sign in to vote

  Correction.

  OCS is operating correctly.

  The sequence is the PSTN user is closing the call.

  No BYE is being sent by the gateway to OCS to signal this.

  The OCS user then tries to close the call manually but it is the gateway responding <----SIP/2.0 481 Call Leg/Transaction Does Not Exist

   

  This would seem to indicate that the gateway is aware that the PSTN call has been closed.

  It is not sending a bye to the mdiation server but it is tearing down the call-id etc

  When the OCS user finally manually hangs up the mediation server is sending a bye to the gateway which is responsing with "Call Leg/Transaction Does Not" because it has already torn it down without telling OCS.

   

  I am talking to dialogic about this and will post a response.

   

  Wednesday, December 19, 2007 1:11 PM
• 

  

  0

  Sign in to vote

  The issue is caused by me having two NICS in the server both on the same network (subnet).

  Using two nics on different networks or a single NIC\IP stops the issue occuring.

   

  More Detail

  "

  When OCS initiates a voice call out to the PSTN via a gateway it sends a SIP invite in the form

  CONTACT: <sip:Mediationservername.com:5060;transport=Tcp;maddr=192.168.10.1;ms-opaque=ba203bbfbffd3f5f>

  When a gateway sends a bye,it will attempt to send it to the IP address specified in the maddr in the Contact address at port 5060.

  The MS best practises is to use two NICs on a mediation server. One card faces the gateway; the second card faces the Communications Server 2007 server that acts as the Mediation Server’s internal next hop. The IP address that you select from the Communications Server listening IP address must match the address that is returned by a DNS query on the Mediation Server’s FQDN. (It is possible to configure both edges on a single adapter card, but not recommended)

  However if both NICs are on the same subnet the mediation server uses the NIC highest in the Binding Order. Typically this is the Communications Server listening IP and it will reject traffic from the gateway. (Gateway attempts to open a TCP socket to port 5060 at the IP address specified in the maddr in the Contact address, but the destination is responds with an RST/ACK, which rejects the socket open. Because of this gateway cannot send the BYE)

   

  Effectively when it comes to production the NICs should be on different networks. (The Mediation server <=> gateway should be a closed vlan or potentially a cross over cable)

  For now the issues have been resolved by configuring both edges on a single adapter card

  "

  Monday, January 14, 2008 10:51 AM
• 

  

  0

  Sign in to vote

  I have the same problem. I have a 2800 series Cisco Router with Voice Card. Single ISDN PRI. Mediation Server. Mediation Server is on same subnet as OCS farm and Gateway is on subnet that handles the PSTN demarks (phone room which is across campus from the datacenter). So i know they are on different subnets but i have tested in both a dual nic (different subnets) and single nic on different subnets and nether config works.

   

  Some notes:

   

  I have used a Dialogic 2000 GW in same single NIC config and it does not do it.

   

  The problem looks like the Cicso is sending a CANCEL and not a BYE since the call is never picked up and connected. This would make sense in this scenerio.

   

  I plan to do packet captures on the dialogic and see if it is sending a BYE instead of a CANCEL.

   

  My question is, are there people using Cisco GW's and if so did you have to change any SIP settings on the router that were not default.

   

  With the Dialogic working just fine and the Cisco (using exact same network details) experiences this behaviour.

   

  We are mainly a Cisco shop and i have would like to stay standardized. The cisco actually seems to sound better too.

   

  Any Cisco Gateway users out there and if so what kind of config did you use to get it working? This is the only issue so far with the Gateway. Everything else is great.

   

  Rick Phillips

  University of Kentucky

   

   

  Thursday, February 21, 2008 3:00 PM
• 

  

  0

  Sign in to vote

  Microsoft won't support this scenario at the moment however we have it working in a lab & I know Microsoft are looking into it also.

   

  Have a look at. (This document isn't great but it is a start)

  http://www.cisco.com/en/US/solutions/ns340/ns414/ns728/networking_solutions_products_genericcontent0900aecd805bd0e4.html

  See Microsoft Office Communication Server 2007 version RTM to Cisco IOS Voice Gateway using SIP with E1 ISDN

   

  What we have is a 2811 router in the lab. It did not work until we upgraded the IOS to 12.4(11)T4

  We are actually using it as an IP to IP gateway talking to CCM 4.13.

   

  We have done quite a few tests and overall it seems to work well. We have only one slight glitch with

  Cisco phone diverts a call so the end result is two ocs users being connected after about 60 seconds the call is dropped. All the other call transfers, holds, conference etc seem to work well. However if you are not using CCM you won't come up against this.

   

  Hope this helps.

   

  Regards

  Alistair

   

  Thursday, February 21, 2008 5:14 PM
• 

  

  0

  Sign in to vote

   

  Hi Guys,

   

  Can i check with you your integration design?

   

  pstn ------>router---------> CM 4

     \---e1--> Dialogic GW----sip---> Mediation

   

   

  Is your design like this? Is it possible for you to post your router configuration? I can't seems to get the connection working between the router and the Dialogic GW.

   

  Thanks.

   

  Regards,

  Lee WH

  Wednesday, March 26, 2008 3:23 AM
• 

  

  0

  Sign in to vote

  For our lab we had

  pstn => router 1 => CCM4.13 => router 2 - e1 -> Dialogic => Mediation

  For production I left it to the Voice Team

  On CCM. the gateway will be a standard MGCP type with the following changes

  Interface information

   

  PRI Protocol Type = PRI QSIG E1

  Protocol Side = Network, as the TMG will be set to terminal as default

  PCM type = A-law

   

  Key parts of router config _ this is an old version (We are trying to use the router as an IP to IP gateway _ ie no dialogic required with some success.)

   

   

  card type e1 1 1
  !
  no aaa new-model
  !
  resource policy
  !
  network-clock-participate slot 1
  network-clock-select 1 E1 1/0
  ip subnet-zero
  !
  ip cef
  !
  no ip domain lookup
  isdn switch-type primary-4ess
  isdn voice-call-failure 0
  isdn gateway-max-interworking
  !
  voice-card 0
   dspfarm
  !
  voice-card 1
   dspfarm
  !
  voice rtp send-recv
  !
  voice service voip
  !
  username all
  !
  controller E1 1/0
   pri-group timeslots 1-31 service mgcp
  !
  interface Serial1/0:15
   no ip address
   isdn switch-type primary-qsig
   isdn overlap-receiving
   isdn protocol-emulate network
   isdn incoming-voice voice
   isdn T310 120000
   isdn bind-l3 ccm-manager
   no cdp enable
  !
  control-plane
  !
  voice-port 1/0:15
   cptone C1
  !
  ccm-manager fallback-mgcp
  ccm-manager mgcp
  ccm-manager music-on-hold
  ccm-manager config server xxxxxxxx
  ccm-manager config
  ccm-manager download-tones
  !
  mgcp
  mgcp call-agent xxxxxxxxxxx 2427 service-type mgcp version 0.1
  mgcp dtmf-relay voip codec all mode out-of-band
  mgcp rtp unreachable timeout 1000 action notify
  mgcp modem passthrough voip mode nse
  mgcp package-capability rtp-package
  no mgcp package-capability res-package
  mgcp package-capability sst-package
  no mgcp package-capability fxr-package
  mgcp package-capability pre-package
  no mgcp timer receive-rtcp
  mgcp sdp simple
  mgcp rtp payload-type g726r16 static
  !
  mgcp profile default
  !
  dial-peer cor custom
  !
  dial-peer voice 9 pots
   destination-pattern 33..
   progress_ind alert enable 8
   progress_ind progress enable 8
   progress_ind connect enable 8
   direct-inward-dial
   forward-digits all
  !
  dial-peer voice 3300 voip
   destination-pattern 330[0-5]
   progress_ind setup enable 3
   progress_ind alert enable 8
   progress_ind progress enable 8
   progress_ind connect enable 8
   session target ipv4:xxxxxxxxxxxxxxxxxx
   incoming called-number .
   dtmf-relay h245-alphanumeric
   codec g711ulaw
   no vad
  !
  dial-peer voice 10 pots
   destination-pattern 330[5-9]
   port 1/0:15
   forward-digits all
   prefix +44xxxxxx
  !
  scheduler allocate 20000 1000
  ntp master
  !
  end
  

   

   

  Wednesday, March 26, 2008 1:28 PM
• 

  

  0

  Sign in to vote

  Hi Alistair,

   

  Thanks a lot.

  I will work on this and let you know the result.

   

  rgds,

  lee

  Wednesday, March 26, 2008 1:39 PM
• 

  

  0

  Sign in to vote

  Hi AlistairK,

   

  I have a mediation server in our lab with both Nics on the same subnet and it works fine. I think a key to it is to ensure that the Nic that connects to the OCS front end is registered in DNS and the other nic that is used for talking to the PSTN gateway is NOT regisitered in DNS. This can be done in the TCP/IP properties under the Nic configuration in Windows. Just unselect the register in DNS check box for the PSTN nic. I know its a little late but it may help anyway.

   

  Cheers

  Chris

   

  Wednesday, March 26, 2008 2:31 PM
• 

  

  0

  Sign in to vote

  In the scenario I had, problems would only occur under certain instances. Like you I set the NIC to not register in DNS.

  The issue was that OCS would use the NIC\IP highest in the binding order to connect to the dialogic gateway.

  When the PSTN user terminated the call the dialogic gateway would notify the IP address which had initated the connection.

  If this IP was not the gateway listening IP the mediation server would reject the connection.

   

   

  Wednesday, March 26, 2008 2:49 PM
• 

  

  0

  Sign in to vote

  Hi Alistair,

   

  May I know what cable are you using for connection from router 2 to dialogic? I am using a RJ45 straight cable.

   

  Rgds,

  Lee WH

  Thursday, March 27, 2008 1:21 AM
• 

  

  0

  Sign in to vote

  Ahh you will need a special cable.

  RJ45 cable but pin out for E1\T1 CSU/DSU Cross-Over Pinout

  http://www.cisco.com/en/US/products/hw/routers/ps233/products_tech_note09186a00800a3f09.shtml

   

  Thursday, March 27, 2008 9:02 AM
• 

  

  0

  Sign in to vote

  i have it working after changing the cable. it was the cable that causing the problem all this while. thanks a lot.

   

  Friday, March 28, 2008 11:27 AM