Restored v18 backup, Trunks refuse to connect to VoIP provider

[per.eriksson_1]

Hi

After restore v18 backup to v18.0-9 debian server I get SBCs and Webmeeting to connect fine but it seems that the provider trunks are dead. I can't see anything happening on port 5060.
I have deleted the trunks and created again but....dead...it doesn't even try....

All services are running and i don't see anything in event logs, dashboard is all green and firewall test is all ok. i dont use 5090 internally. Only for the provider trunks so its not part of the firewall test.

No logging shows up on the trunk lines...





What have I forgot???

BR
Per

 

[AWS2P]

is it a registred Trunk or IP Based ? if ip based is your Public IP always the same ?
is your local IP also static and the same it was until now ?

no change on your network? no Firewall rules change?

Have you got support from your SIP PRovider ?

 

[per.eriksson_1]

Registered Trunk
Entire network setup has changed from PFsense router to AWS VMWare NSX.
Local static IP on a new subnet compared to the original server (the reason for reinstallation from backup)
New public IP and entire path to internet is new.
Support from the provider is limited to "we do not see any packages from your IP" which I have already figured out as I get no visible 5060 taffic from 3cx.
If I do UPD ping via port 5060 from the 3cx server to another external Linux machine, I see the pings nicely in the other end. If I then put that test servers IP as the address for the trunk then nothing happes on 5060 in the other end.
External UPD 5060 ping reaches the server through the firewall as well so the firewall is OK.

3cx is straigh up silent on the provider trunks. Not a lifesign....

 

[SteveITS]

Is DNS functioning on the 3CX server? What do the activity log and event log show?

 

[per.eriksson_1]

Its a DNS problem. Direct IP on the trunk works....

The OS can ping the DNS address to the VoIP provider but not from within the 3cx application....what the heck...

 

[AWS2P]

have you run the 3CX firewall checker? is it green with no problem for 5060?

your screen copy shows Fw as green but is it a test done now in latest install?

You think it's DNS related so what is set as DNS server on your VM ? 8.8.8.8.?

 

[per.eriksson_1]

I see no errors related to the trunks nor 5060 in the logs when i use the DNS addresses.

All fine now when I put the trunks with IP instead of DNS addresses. No issues at all.

I have my own internal DNS servers (to be able to do the required DNS loopback for the FQDN).
From the OS I have no issue to ping DNS addresses.

Is 3cx performing something spooky to do its own DNS resolution that is not related to how the OS perform DNS resolution?

 

[AWS2P]

From LAN when you nslookup your FQDN is it your private local IP you get ?
Doing the same for outside, nslookup FQDN 8.8.8.8 ? do you get your actual public IP ?

what's strange is SBCs are able to connect to the pbx and they obviously use DNS resolution to it.but they are using 5090 instead of 5060.
Are you sure you have no firewall rules blocking outside LAN things with 5060? or old rules with previous IP remaining active and polluting DNS resolution ?