In order to take advantage of the new phone directory module we upgraded from 2.5 to 2.8. Everything went well. However our clients are now reporting intermittent but consistent upstream call quality issues (I can hear them but they can’t hear me). As well it only happens on outbound calls.
We use a major SIP trunking provider and have had no problems previous to the upgrade. I have looked at all the sip configs files and hardly anything has changed between versions of Free-PBX. Also our Asterisk version has not changed, it is 1.4.27.
Our link is good as we are colo’d at a good data center, and we have done ping plots to all the parties involved.
My question is, is there some setting buried somewhere that may have caused this problem. Any feedback would be appreciated.
the only thing that really could be related is if the configuration was being changed. Furthermore, a configuration change would not result in intermittent audio issues. They would either be always or never.
Since you Asterisk version did not change (which also would not likely be related to intermittent issues), you should be looking closer at your firewall and likely working closer with your carrier to determine why this is happening.
As a long shot, you may want to check if the codecs being configured (which may come from default codecs) are the same and in the same priority order as before.
Thanks for the quick response. We kind of thought the same thing, it was just very coincidental, we upgraded on the weekend and then had a spike of complaints since Monday. However we have no hard evidence that FreePbx is the cause. We are working with the carrier and doing the tedious work of going over the logs.
I was hoping we where going get lucky and you where going tell us a bout some fundamental change in the way the SIP config was handled between versions.
To answer your questions though:
The machines have no firewall in front of them (they have been secured)
The codec order is ulaw, alaw, and thats it.
I am not going to go into it in an open forum though. A firewall layer represents a point of failure we wish to eliminate.
We do however employ real time sniffing of ALL packets traversing our network, and we in turn have st up alarms to deal with certain traffic scenarios. We can also divert/null route/apply ACL’s at our routing layer.
Contact me directly if you want to discuss further.
We’re making good headway with our carrier and I will post final reselution when we get it solved.