Declined calls - fine after reboot

[dndx]

Hello!
For about two weeks, I am using 3CX (4SC, Standard) on a small Debian Server with the provided ISO. While I am very happy with the configuration and features, the stability has been... insufficient.

At first, I tested 3CX on my personal computer to see if I am able to configure everything before actually buying hardware. I used the free version, but after some reboots I noticed that I can no longer call anyone or be called - every phone said "declined", from outwards I get the "temporarily unavailable" message. In the frontend log incoming calls appear as missed, both internal and external calls. After another reboot, everything was working instantly again, so I didn't give it any more thoughts.

Unfortunately, the problems reoccur with what I felt to be like every 3rd reboot. I even reinstalled everything twice, set it up from scratch (no backups used) to eliminate one-time installation errors. The problems are the exact same: If the machine is restarts, there is a high chance that everything stops working after the boot is completed. Internal calls are affected as well as external calls. My two SIP Trunks as well as all the phones show up as registered. The Yealink T46S are also showing green BLFs for the extensions during any time.

Does someone know how this problem can be solved? I would like to provide additional log files, however, I do not know where to find them so maybe someone can show me under which location they are saved so I can provide additional information.

Thank you!

 

[Brian Cross]

Sounds like an issue with the networking on the server if I am understanding you correctly... Static IP? when it stops working what is the status of eth0?

 

[dndx]

Yes, I use a static IP with my server now, on the testing computer I used DHCP.
Networking works fine at any time. Every phone logs in nicely, the control panel is available and internet access works also regardless of the issue.

 

[Brian Cross]

So after a few reboots, trunks show connected, phones show connected with BLF buttons working, but you can't dial in or out? Just making sure I understand what is happening.

Firewall checker passes all checks as well?

 

[nb]

Seems it is working now..then please update the title of the post or mark as solved.
Because when an ip is insufficiently static (it changes all the time) then the pbx gets it's wheels replaced every time a reboot takes place. You won't be going very far with this configuration.
And check whether your DHCP can do reservations because DHCP reservations is more stable than setting a static fixed ip on a host. Its like you invent or choose your own postal address. You might make an address that conflicts with another address or some other device will take the ip address of the pbx. You should let your address come from the service entitled to send addresses - the DHCP.

 

[dndx]

Thanks for your replies!

No, the issue is not resolved. To test the system, I used my PC first (which uses DHCP reservations) and the issue occurred, and after installing 3CX on the server with a static IP the issue occurs as well. I am also certain that the IP address is not used by any other client

So after a few reboots, trunks show connected, phones show connected with BLF buttons working, but you can't dial in or out? Just making sure I understand what is happening.

Exactly, this is what's happening.

 

[3CXDude]

I've experienced something similar recently, recently on 2 seperate systems (15 & 15.5 windows)

Everything looks fine in the management console but the trunks are not registered. No email notifications it's just like the trunk freezes and the first we hear of it is the customer calling us to complain.

Only thing that resolves is a reboot.

 

[dndx]

My trunks are fine, they and everything else is registered. Because I also can't call internal extensions then, your problem might be a different one.

 

[nb]

DNDx: If you cannot call internal extensions you have a problem of a different dimension. I think your ip address is not stable or not static.
Also check memory and cpu usage. Maybe there is a race condition and things are hogged up.

3CXDude: Before you reboot check that all services are up. Try and reach out to me when it happens - let me just have a look.

 

[dndx]

Thanks again for your reply!

Trust me, the network connections are fine Setting up networks is part of my job so I would say everything is good on that side. The IP of the 3CX Server is static (the phones IP is static as well)

Both memory and CPU are very low too, it runs on an Intel NUC with 4GB RAM and an Intel Quad Core J3455 with 4x2.3Ghz. Memory usage is currently at 0.9GB/4GB and I have not seen a CPU usage of more than 5% during normal operation, linux load average over the past 15 Minutes is 0.01. Currently it works fine, but because I also checked the interface when the issue occurs including the memory and cpu usage, it can't be very different, otherwise I would have noticed that.

Do you know under which path I can find a detailed log? I think think this would really help.

 

[nb]

Make a wireshark capture from oner internal extension to the other. When A calls B A sends invite to server and server sends to b.
Can you check if server is sending this to B or not?
Can you check if Server is accepting A's invite or not?
Server activity log in verbose should show you more details of what is happening.

 

[dndx]

Is it possible that the server activity log is deleted with every reboot?

I tried to get as much log information as possible but it seems like everything was deleted, the only log that I got from the Logs Backup folder was the 3CXDialer.log which looks like this, it is from when the system did not work:

17/08/30 16:28:49.327|200002| Trc|75|0009|: Ins: Call(30) [Dialing] DN=10001 IP=10001 EP=<number_that_is_calling>
Attached: [sip_displayname=<number_that_is_calling>;lookup_displayname=;chid=0000015E338B6E27_30;prevCall=0;prevLeg=0;extnumber=<number_that_is_calling>;devcontact=sip:<SIP-TRUNK>;caller_name=<number_that_is_calling>]
17/08/30 16:28:49.328|200002| Trc|75|0009|: Del: Call(30) [Dialing] DN=10001 IP=(null) EP=<number_that_is_calling>
Attached: [sip_displayname=<number_that_is_calling>;lookup_displayname=;chid=0000015E338B6E27_30;prevCall=0;prevLeg=0;extnumber=<number_that_is_calling>;devcontact=sip:<SIP-TRUNK>;caller_name=<number_that_is_calling>]

I will try to collect log information before a reboot next time, and also try wireshark. Right now I can't do that, I will on the weekend.

 

[nb]

But I asked you to make a Wireshark capture. This you can do now you dont need to wait.
Just do it let me see what is happening.. takes you 2 seconds. With the pcap function in 3CX, you can do this literally in 2 secs.

 

[dndx]

Shouldn't I do a wireshark capture when I have the problem with the non-working calls? Right now everything works fine. To trigger the issue, I would have to reboot the system a couple times but right now the phones are in use.

 

[nb]

Yes - i thought you had the problem now.
So yes agreed. When you have a problem do it. If you have a problem without rebooting, its perfect. Do not reboot when you get the problem.

 

[3CXDude]

@nickybrg

All services are up, have a ticket open with support on it. Have verbose/shark running waiting on it to happen again to send over to them.

Will let you know.

 

[nb]

Dont leave wshark running.. You dont need. And it is dangerous for the server. Stop it and leave verbose logs only.
When it happens you inform us.