Dead Air - Internal and Outbound Calls

[GreasyGopher]

Hello,

I am currently experiencing an issue and was wondering if anyone has run into this before. The issue is internal calls within the office and outbound calls result in dead air. The issue seems to be intermittent. When it stops working it wont work until we reboot the services on the 3CX server. Then after would could be hours or days it will start resulting in dead air calls again. In bound calls work just fine. This seems to have started happening since the last update to 3CX.

One of our other environments has the exact same 3CX setup and settings. Same firewall settings. No custom VOIP rules. Consistent NAT enabled. SIP transformations disabled. They have no issues what so ever.

-Hosted 3CX server 15.5.0
-Yea Link T42S phones
-Sonicwall SOHO
-Firewall checker passes.

Problem environment is as follows:

-Hosted 3CX server 15.5.0
-Yea Link T42S phones
-Sonicwall TZ215
-Firewall checker fails

• resolving 'stun-us.3cx.com'... done
• resolving 'stun2.3cx.com'... done
• resolving 'stun3.3cx.com'... done
• resolving 'sip-alg-detector.3cx.com'... done
• testing 3CX SIP Server... done
  • stopping service... done
  • detecting SIP ALG... not detected
  • testing port 5060... done
  • starting service... done
• testing 3CX Tunneling Proxy... done
  • stopping service... done
  • testing port 5090... done
  • starting service... done
• testing 3CX Media Server... failed (How to resolve?)
  • stopping service... done
  • testing ports [9000..9398]... done
  • testing ports [10600..10998]... failed (How to resolve?)
  • Cone testing failed
  • starting service... done

Any suggestions would be appreciated. I shortened the port checker log to fit the post.

 

[leejor]

testing ports [10600..10998]... failed (How to resolve?)

This would probably be the issue, for external calls. There have been a number of posts about this lately (if you look through the forum), as the port range has been expanded recently. You will need to be sure that the additional ports are open in your firewall/forwarding.

Who is you provider? There have been posts specifically about AT&T settings.

https://www.3cx.com/community/threads/occasional-one-way-audio.59653/

 

[GreasyGopher]

I have been entertaining this idea, but what confuses me is why our other environment doesn't need this firewall/forwarding to function properly and have a nearly identical setup there.

 

[leejor]

Does the Firewall Checker pass or fail, at the other sites? If it passes, is there is a difference, somewhere.

 

[Alphabetic]

Have you got PBX delivers audio enabled on extensions?

 

[cobaltit]

I have been entertaining this idea, but what confuses me is why our other environment doesn't need this firewall/forwarding to function properly and have a nearly identical setup there

I mean the firewall checker is staring you right in the face.....

Yes you need to update your firewall rules in both places. Both your SIP trunking and remote extensions will use that expanded range.

 

[cobaltit]

As far as why it's a problem for this setup and not the other ones is also self-evident in your statement of 'nearly identical'. I would wager the difference is the other site is not on Update 6/SP6 where the port range was changed or the other site has less call volume so the system takes longer to creep up into the range you currently aren't allowing.

 

[GreasyGopher]

Heres an update on the situtation.

1. PBX delivers audio enabled on extensions is enabled already.
2. The filewall checker issue has been resolved and the checker passes but we are still experiencing dead air. On top of this we actually found out that all of our implementations of this phone system are experiencing this but only some offices are noticing it more or being more vocal about it.

We are using Stun mode and 3CX is recommending that each phone has a static IP and the firewall to have static port forwarding with a range of 11 ports for each phone. They are yealink phones.

Has anyone had to do this for any of their setups?

 

[Alphabetic]

Hi

Yes I’ve had to do this on my STUN customers. Though not all saw a problem (smaller setup) but I did it for them anyway.

Customer noticed improved performance (minimal delay) since implementing this changes suggested to you.

I would recommend purchasing a Pi and using an SBC