- Joined
- Oct 5, 2009
- Messages
- 154
- Reaction score
- 0
Hi all,
A bit of background on this topic. We provide our clients with VoIP systems and host them in house with all sites phones connecting to their hosted dedicated 3CX server via hardware VPN tunnels using a mix of 10Mb Fibre lines or dedicated ADSL circuits.
Prior to 3CX we used an alternate VoIP system that had feature limitations but was rock solid with call quality (the actual SIP gateway element of the previous solution was a dedicated hardware device - Ingate SIPerator firewall).
We changed to 3CX last year after a successful 3CX trial with our internal system and now have several 3CX based installs in use for clients. On each site they have all mentioned experiencing intermittent issues with delay on VoIP calls. This is not something we experience in house on our system (but we are on the same LAN as our system) or expected to encounter for our client systems.
We have looked at various possible causes from the comms in use on the clients site, to the SIP trunk provider in use, but in each case nothing appears to be the cause and in any case it's the same comms and SIP trunk provider/LAN topology that were used in our previous non 3CX implementations that didnt experience these issues.
We then began to think that it must be 3CX that's the cause and that the software SIP/Media gateway is adding additional overhead to the media processing and this is causing the delay, but if this was the case I would expect the delay to happen on each and every call (which it doesn't)?
My question is - is anyone else out there using 3CX with remote phones connected via VPN and experienced/troubleshooted similar issues?
My experience of troubleshooting VoIP issues of this nature is limited but we have taken call recordings of "delayed" calls from the server and cannot hear any delays, yet we can hear the participants in the call commenting on the delay in the actual call audio itself.
We have also taken wireshark captures on the server and can see the 2 legs of the call in wireshark but I've not used wireshark to this level of depth before and am not sure which legs of the conversations shown will make up the audio received by each party - also, will the capture taken on the server necessarily reflect any delays experienced at the client site?
I presume to get a true picture I also need a capture taken off a mirrored port onsite which can then be compared to the server based trace to see if the delay is being added by the VPN?
It's a massively frustrating problem to get to the bottom of and it's not helped by it's intermittent nature. I could understand if poor comms were to blame but I would expect this would manifest itself in the way of dropped rather than delayed audio? We also used the same comms with the different system before 3CX and didnt experience these issues.
Can anyone offer any advice/assistance with regards for what I should be looking at in traces/3CX logs?
We have an engineer going to a client site tomorrow to do some local captures so hopefully that will provide a clearer picture.
I am championing 3CX within our company but I will not be able to progress much further if I can't prove this problem is not caused by 3CX.
Thanks
Paul
A bit of background on this topic. We provide our clients with VoIP systems and host them in house with all sites phones connecting to their hosted dedicated 3CX server via hardware VPN tunnels using a mix of 10Mb Fibre lines or dedicated ADSL circuits.
Prior to 3CX we used an alternate VoIP system that had feature limitations but was rock solid with call quality (the actual SIP gateway element of the previous solution was a dedicated hardware device - Ingate SIPerator firewall).
We changed to 3CX last year after a successful 3CX trial with our internal system and now have several 3CX based installs in use for clients. On each site they have all mentioned experiencing intermittent issues with delay on VoIP calls. This is not something we experience in house on our system (but we are on the same LAN as our system) or expected to encounter for our client systems.
We have looked at various possible causes from the comms in use on the clients site, to the SIP trunk provider in use, but in each case nothing appears to be the cause and in any case it's the same comms and SIP trunk provider/LAN topology that were used in our previous non 3CX implementations that didnt experience these issues.
We then began to think that it must be 3CX that's the cause and that the software SIP/Media gateway is adding additional overhead to the media processing and this is causing the delay, but if this was the case I would expect the delay to happen on each and every call (which it doesn't)?
My question is - is anyone else out there using 3CX with remote phones connected via VPN and experienced/troubleshooted similar issues?
My experience of troubleshooting VoIP issues of this nature is limited but we have taken call recordings of "delayed" calls from the server and cannot hear any delays, yet we can hear the participants in the call commenting on the delay in the actual call audio itself.
We have also taken wireshark captures on the server and can see the 2 legs of the call in wireshark but I've not used wireshark to this level of depth before and am not sure which legs of the conversations shown will make up the audio received by each party - also, will the capture taken on the server necessarily reflect any delays experienced at the client site?
I presume to get a true picture I also need a capture taken off a mirrored port onsite which can then be compared to the server based trace to see if the delay is being added by the VPN?
It's a massively frustrating problem to get to the bottom of and it's not helped by it's intermittent nature. I could understand if poor comms were to blame but I would expect this would manifest itself in the way of dropped rather than delayed audio? We also used the same comms with the different system before 3CX and didnt experience these issues.
Can anyone offer any advice/assistance with regards for what I should be looking at in traces/3CX logs?
We have an engineer going to a client site tomorrow to do some local captures so hopefully that will provide a clearer picture.
I am championing 3CX within our company but I will not be able to progress much further if I can't prove this problem is not caused by 3CX.
Thanks
Paul