Internal P2P call route to edge pool sometime not working

There have two edge server on one edge pool, lync to lync call the call is going to peer to peer call but p2p subnet is block and it will go to edge server for the audio.

if the the call is go to edge 1 Public IP and go through edge 2 Public IP , there will have problem.

there edge server is doing NAT so the WAN to WAN interface is block.

22; reason="Call failed to establish due to a media connectivity failure when both endpoints are internal"; UserType="Callee";MediaType="audio"; NetworkErr="no error"; ErrTime="0"; RTPSeq="0"; SeqDelta="0"; RTPTime="0"; RTCPTime="0"; TransptRecvErr="0x0"; RecvErrTime="0"; TransptSendErr="0x80004005"; SendErrTime="3649479797252"; InterfacesStall="0x0"; InterfacesConnCheck="0x0"; MrDnsE="av.ABCpacific.com"; MrResE="0"; MrDnsI="edgepool1.ABC.local"; MrResI="1"; MrDnsCacheReadAttempt="0"; ICEWarn="0x40003a0"; ICEWarnEx="0x0"; LocalSite="192.168.80.55:23222"; LocalMR="Public IP 1:54459"; RemoteSite="192.168.100.69:32474"; RemoteMR="Public IP 2:57310"; PortRange="1025:65000"; LocalMRTCPPort="54210"; RemoteMRTCPPort="57310"; LocalLocation="2"; RemoteLocation="2"; FederationType="0"; NetworkName="ABC.local"; Interfaces="0x4"; BaseInterface="0x4"; BaseAddress="192.168.80.55:28392"; IceRole="1"; RtpRtcpMux="1"; MediaDllVersion="5.0.8687.139"; MrDnsE="av.ABCdomain.com"; MrResE="0"; MrDnsI=" edgepool1.ABC.local"; MrResI="1"; MrDnsCacheReadAttempt="0"

Any advise?

Thanks.

August 25th, 2015 9:13am

Hi Jo

It looks as though the media is trying to internal "Call failed to establish due to a media connectivity failure when both endpoints are internal"

Also the ICEWarn Flag confirms this, translating to

  • Local connectivity failed
  • UDP TURN server connectivity failed
  • TCP TURN server connectivity failed

Check your firewalls, NATs and routes from this subnet

thanks

Free Windows Admin Tool Kit Click here and download it now
August 25th, 2015 9:53am

This subnet to edge server is work.

Add static route on edge server for every subnet.

How to force internal traffic to go Edge internal face and don't go out WAN public.

If I stop one of edge server the Lync to Lync call and video is work, because the call is go through one edge WAN interface and back to same WAN IP.

Please check this message is Public IP 1 to Public IP 2

LocalMR="Public IP 1:54459"; RemoteSite="192.168.100.69:32474"; RemoteMR="Public IP 2:57310";

Thanks.

  • Edited by Jo Tse Tuesday, August 25, 2015 11:00 AM change
August 25th, 2015 10:02am

I found if edge pool have two server, it need enable pairpinning on firewall for public A/V ip communicate from Edge Server A to Edge server B.
Free Windows Admin Tool Kit Click here and download it now
August 26th, 2015 5:01am

Does anyone know there have any method to do? 

force the Edge server A traffic don't go through public ip to another Edge server B public ip.

August 26th, 2015 12:27pm

As per the client logic in order to establish the call, both clients exchange their IP addresses to find out best possible IP's from both ends to get the call established. 

and in your case even though the clients are reachable to each other using the internal IP so the media is trying to flow internally, but to configurations on the switch does not allows the media to flow between the clients. 

Your requirement to route the voice/audio via the edge server even though the IP address of caller and callee are internally rotatable, is basically defying the client logic.   

you may try to configure the switch to allow only VOIP packets based on certain ports only by making use of QOS and DSCP marking as explained in

http://blog.schertz.name/2014/10/configuring-qos-for-lync-ip-phones/

https://technet.microsoft.com/en-us/library/gg405409(v=ocs.15).aspx

 

 

Free Windows Admin Tool Kit Click here and download it now
August 30th, 2015 1:02am

This topic is archived. No further replies will be accepted.

Other recent topics Other recent topics