On 8/19/2010 9:26 PM, Martin Steinmann wrote: >> I went through the requirements in detail on this list, and I never >> could find a plan that would work. >> There were 2 huge issues. >> 1)I never understood a way I could have a seamless transition to an fxo >> for outbound calls if the connection to the central office went down. > Audiocodes GWs can do this. It became stable starting firmware release 5.8 > and it is supported in sipXecs 4.2. > http://wiki.sipfoundry.org/display/xecsuserV4r2/AudioCodes+6.00+Stand-Alone+ > Survivability > I started with 4.0.2, so perhaps this is something that may have worked, but wasn't available at the time. I am using Audiocodes MP 114s for the backup gateway. My outbound traffic at the remote location needs to be SIP. It seems I have to have an SBC of some sort at the remote office (discussed more below) > A second option would be to configure the local gateway as an emergency > gateway in the phones. For Polycom phones you can do this through the UI on > the phone screen adding a line, then go to tab Dial Plan. If the main route > to sipXecs cannot be found, the phone uses the 2nd one, which points at the > local GW. You can configure this on the phone group level as well. > Is it possible for the user to dial any call exactly the same on the handset with no knowledge of whether or not their primary SIP connection is down? >> 2)I never understood a way I could make the call traffic go in and out >> the local mpls connection at the remote office without putting a local >> sipx device of some sort at the remote site. My understanding was if >> sipx supported media release, this would have been possible. > Explain 'media release'. sipXecs separates media from signaling and local > media stays local. > Maybe that isn't the correct term. That is what Verizon called it. In this case, it is the ability for the Sipx server to drop out of the picture for RTP traffic. Asterisk can do this. I was under the impression Sipx couldn't do this by design since it was a proxy. I need the sip RTP traffic for a remote site to go in/out their local MPLS connection. I can't have it coming back through our corporate office. Here is a description: http://www.voip-info.org/wiki/view/Asterisk+sip+directrtpsetup All my endpoints can access each other. There is no IP NAT involved. Verizon sends calls to our servers by their actual IP and could communicate directly with the handsets if needed. This is what my VoIP carrier wanted to do. Sip signaling would go through the corporate office, but RTP traffic would go directly to and from the remote site.
>> I wanted to do what you describe, but I was unable to find a way to >> make it work for us. I'll be glad to keep discussing if you think there >> is a way. This was my initial conversation on the topic: http://www.mail-archive.com/sipx-users@list.sipfoundry.org/msg10781.html I would love to find a way to make it work. I hit a roadblock with every scenario I worked on. > --martin > >> -----Original Message----- >> From: "Martin Steinmann"<mstei...@gmail.com> >> Date: Thu, 19 Aug 2010 21:49:43 >> To:<mkitchin.pub...@gmail.com>; 'Michael >> Scheidell'<michael.scheid...@secnap.com>;<sipx- >> us...@list.sipfoundry.org> >> Subject: RE: [sipx-users] port 5060/ port 5080, proxy why? >> >> Why not a centralized deployment with only phones and optional gateways >> in >> the remote office? Having to manage 110 small ITX boxes does not >> sound >> pretty. >> --martin >> >>> -----Original Message----- >>> From: Matthew Kitchin (Public) [mailto:mkitchin.pub...@gmail.com] >>> Sent: Thursday, August 19, 2010 9:43 PM >>> To: Martin Steinmann; 'Michael Scheidell'; sipx- >>> us...@list.sipfoundry.org >>> Subject: Re: [sipx-users] port 5060/ port 5080, proxy why? >>> >>> Using 2 hosts (sipxbridge on a dedicated one) was the other option we >>> looked at. I didn't do it for 2 reasons. I was a total novice and >>> wanted to keep things simple. And, our corporate office was the model >>> we would follow at our 110 small remote locations. We wanted to do >>> small mini itx (on a bberry, I think that is what they are called) >>> boxes at the small sites, and adding a second box wasn't practical. >>> -----Original Message----- >>> From: "Martin Steinmann"<mstei...@gmail.com> >>> Sender: sipx-users-boun...@list.sipfoundry.org >>> Date: Thu, 19 Aug 2010 21:30:34 >>> To: 'Michael Scheidell'<michael.scheid...@secnap.com>;<sipx- >>> us...@list.sipfoundry.org> >>> Subject: Re: [sipx-users] port 5060/ port 5080, proxy why? >>> >>> _______________________________________________ >>> sipx-users mailing list >>> sipx-users@list.sipfoundry.org >>> List Archive: http://list.sipfoundry.org/archive/sipx-users/ > _______________________________________________ sipx-users mailing list sipx-users@list.sipfoundry.org List Archive: http://list.sipfoundry.org/archive/sipx-users/
