V20, Event "Your local IP x.x.x.x specified for the 3CX system, has been changed to x.x.x.x - Please adjust configuration." Phones loose connectivity

[iwcb002]

We have a system that was on V18 and would go where the phones went offline and couldn't find 3cx. The system was not locked up, but it couldn't be accessed via the LAN ip in any way. Rebooting the 3cx Linux box would bring everything back. It wasn't locked up, and no logs we could see showed anything.

We upgraded to V20. The same thing has happened a few times now. I did noice an entry in the Event Log though.
Eventlog shows "Your local IP x.x.x.x specified for the 3CX system, has been changed to x.x.x.x - Please adjust configuration."

Ofcourse with IPs in there. The 3cx system has 2 network cards. The first is the "specified" local IP which is the LAN at the site and all the phones are on it, the internet traffic and port forwarding all go to that IP. The second network card is on a different network which is where our SIP Provider's connection is. That second network card has no internet connectivity it can only get to the SIP Servers.

We suspected a power interruption and a weak/failed UPS causing the network switch to loose power, as it happend about the same time over two different weekends. When going onsite to evaluate it the UPS, the UPSs showed 0 events, and on testing the UPS that the swtich was on, it did operate atleast for a short period. We also expected if it was the switch to be able to reproduce it by unpluging the switch, but on unplugging the switch 3cx didn't "change" the LAN ip.

Does anyone have any suggestions why 3cx would "decide" to use the other network card as the local IP?

When this happens, we shutdown the 3cx system and start it back up. Then 3cx is happy and using our proper Local IP, the phones all go online, and we can remotely access it on the LAN. We don't have to change anything.

 

[TheodorosG_3CX]

Have you also configured static routes to avoid this issue?

 

[iwcb002]

There is a static route for the SIP Trunk connection. All traffic to the 10.2xx.x.x range is routed through the second network card that 3cx is saying "has changed to". When it changes, the 10.2.... is working with the SIP Trunk still operating, and you can call in and navigate the Attendant and leave voicemails. Of course no phones work, no apps, softphone, management site etc are accessible in that condition though.

This happens randomly a week or more apart, happening again this morning less than an hour ago, with people in the office.

 

[TheodorosG_3CX]

The PBX is able to detect the ip change and not take action and change it.
Have you ensured the network configuration as per guideline here?

 

[iwcb002]

Is there something that would be expected to trigger 3cx deciding to change to the other network card? Randomly switching after over a week of working as expected is not a behavior that seems typical to me. The second IP is there all the time, it didn't just show up. Also it can not be recreated by simply pulling the cable from 192.x card, at least for only a couple minutes.

 

[TheodorosG_3CX]

The PBX will never request to change the local ip of the server.
You need to check on the local network why it is changing.