For the last 24 hours we have had a strange problem with our 3CX on Debian installation. We use Fanvil X3 phones (version 2.10.1.6836, local on the same VLAN as 3CX server), Windows clients (version 16.3.0.198/16.6.2.89, which can be: local on different VLAN from 3CX, remote via VPN to office where 3CX is installed, or remote via 3CX Tunnel), and the Android clients via 3CX tunnel.
All three different types are continually (generally between 5 and 10 seconds) deregistering and then about a second later reregistering. The result is that although we can place calls to external numbers at any time, the majority of incoming calls, or internal calls, are to lines that deregister before the user picks up, and get diverted to voicemail.
Further information:
- If I go into the 3CX PhoneSystem Terminal from the management console, within a couple of seconds it says "Connection terminated. Please refresh browser window to continue". If I refresh, it says the same within seconds.
- The chat on Android clients is only working when the phone is registered, even though calls can be placed when it is deregistered.
- If it were only the VPN/3CX tunnel clients that were playing up, I would attribute it to unstable internet or something, but the same is happening with ip phones on the same vlan as the server.
- I can ssh into the debian server, and the connection is stable. If I run a ping from the ssh session to the router, to an outside website, or to one of the local phones, no packets are lost. Having said that, after a couple of pings to the phone, the responses seem to take much much longer to come through.
- The server is configured to automatically install updates.
- I have checked the IP blacklist, and there is nothing there that shouldn't be.
- I have checked our IPS, in case it was blocking connections, and it is not. Even if it were, it doesn't explain why the phones on the same VLAN are playing up.
- The only thing that I have changed myself is a few days previous I had started taking backups of the 3CX server (virtualised on Hyper-V) with Veeam Backup and Recovery, but I don't think this should have broken the server in this way.
- I have enabled verbose logging, but can't see anything funny in there (however I'm not a profesisonal, so may have missed something)
If anyone could point me towards what else I should be checking, I would appreciate it. If not I can try setting it up afresh and passing over the settings, but I'd rather not have that hassle if possible.
Thank you!
All three different types are continually (generally between 5 and 10 seconds) deregistering and then about a second later reregistering. The result is that although we can place calls to external numbers at any time, the majority of incoming calls, or internal calls, are to lines that deregister before the user picks up, and get diverted to voicemail.
Further information:
- If I go into the 3CX PhoneSystem Terminal from the management console, within a couple of seconds it says "Connection terminated. Please refresh browser window to continue". If I refresh, it says the same within seconds.
- The chat on Android clients is only working when the phone is registered, even though calls can be placed when it is deregistered.
- If it were only the VPN/3CX tunnel clients that were playing up, I would attribute it to unstable internet or something, but the same is happening with ip phones on the same vlan as the server.
- I can ssh into the debian server, and the connection is stable. If I run a ping from the ssh session to the router, to an outside website, or to one of the local phones, no packets are lost. Having said that, after a couple of pings to the phone, the responses seem to take much much longer to come through.
- The server is configured to automatically install updates.
- I have checked the IP blacklist, and there is nothing there that shouldn't be.
- I have checked our IPS, in case it was blocking connections, and it is not. Even if it were, it doesn't explain why the phones on the same VLAN are playing up.
- The only thing that I have changed myself is a few days previous I had started taking backups of the 3CX server (virtualised on Hyper-V) with Veeam Backup and Recovery, but I don't think this should have broken the server in this way.
- I have enabled verbose logging, but can't see anything funny in there (however I'm not a profesisonal, so may have missed something)
If anyone could point me towards what else I should be checking, I would appreciate it. If not I can try setting it up afresh and passing over the settings, but I'd rather not have that hassle if possible.
Thank you!
