Incomprehensible call loss

[RomainC]

Hi,

I need to set the context to explain the concern:
We are deploying 3CX to a lot of customers and we have a problem that happens only to one of our customers that we can't explain so I turn to the community.
For this customer, they have a switchboard on which 6 people are running, 2 of them every day. The people in charge of the switchboard are complaining about a problem of lost calls:
"When you pick up the call, you hear nothing and the seconds of the call count down to 0. Once at 0, the call window doesn't display anything but seconds that increment this time in the right direction."

After doing some investigation I picked up some info:
problem hard to reproduce because random,
it is not the fact of taking the call at 2 at the same time,
it happens with external calls as well as internal calls and there is no problem once the call is well launched.

Thank you very much in advance, because I am really stuck.

 

[VasilisV_3CX]

Hi @RomainC,

Let's start with gathering a bit more information:

https://www.3cx.com/community/threads/information-to-provide-when-requesting-help.67558/

What app / device are they using that shows a count down?

 

[RomainC]

Hi @VasilisV_3CX,

Sorry I didn't see this thread,

Here are the informations:
3CX Version : 18.0 (Build 314)

Server OS : Linux x64 (Debian GNU/Linux 10 (buster))

Is the 3CX Server Hosted and where? Hosted by us

IP Phone : SENNHEISER SDW3BS

Trunk Provider : OPENIP

Are custom Phone Templates being used: NO

The countdown is visible on the small interface when you receive a call on the desktop version of 3CX Windows v18.

Thank you for your time.

 

[VasilisV_3CX]

Hi @RomainC,

The countdown is visible on the small interface when you receive a call on the desktop version of 3CX Windows v18.

Which client are you using exactly and what version is it? Can you perhaps send me a screenshot?

 

[RomainC]

@VasilisV_3CX,

I use the 3CX Desktop app v18.7.10 for Windows, if this is not the information you wanted, can you tell me where I can find it?

 

[VasilisV_3CX]

Hi @RomainC,

Okay an obvious starting point - Did you uninstall and reinstall the Desktop App?

Is the Debian machine running the PBX running in a virtual environment? If yes, which Hypervisor are you using?

When the countdown ends and the counting starts, does the call continue and you can hear the other party?

 

[RomainC]

Yes no we tried to uninstall and reinstall the application, even more: we upgraded the version of it from v16 to the one currently installed.

No unfortunately I don't have any screenshot to show you, but I confess I would have appreciated it a lot. It happens that it is very difficult for the people who take care of the standard to take a screenshot because it appears only randomly, and it disappears very quickly. So for people who are not necessarily comfortable with the computer tool they do not have time to capture it.

 

[VasilisV_3CX]

Hi @RomainC,

In case you missed my updated post:

Is the Debian machine running the PBX running in a virtual environment? If yes, which Hypervisor are you using?

When the countdown ends and the counting starts, does the call continue and you can hear the other party?

 

[RomainC]

@VasilisV_3CX,

Yes we use your ISO DEBIAN which is already preconfigured with 3CX, and it is a hyper-v cluster.
Once the countdown has reached 0, it starts again in the other direction by incrementing like in a normal call except that there is no call displayed and we still do not hear the person who calls.

 

[VasilisV_3CX]

Hi @RomainC,

Have you ensured that you've installed NTP on the server? Hyper-V has been known to have time drifting issues. Check the bottom of this document: https://www.3cx.com/docs/installing-microsoft-hyper-v/#h.d8s5d1j032hd

 

[RomainC]

Hi @VasilisV_3CX,

I just checked and indeed NTP is already well installed.

 

[VasilisV_3CX]

Hi @RomainC,

To confirm, you mentioned you are using a Hyper-V Cluster, we're talking about a failover cluster configuration using two nodes here? If I've misunderstood feel free to correct me.

Something you may want to investigate is the next time this happens (as it's random), check all the log files you have to see if there is any event that happens at the same time as the random call that counts down. Maybe there is something you can find that connects when your client reports this happening to something in a log file that sheds more light.

 

[RomainC]

@VasilisV_3CX,

Yes a failover cluster composed of 2 servers and a SAN array.

Yes I'll take care of that I'm just waiting for their feedback. But the problem is that when I go through the logs there are no error.

 

[NickD_3CX]

Hi!

Just a follow-up question, you said that these customers were also on V16 and you upgraded them at some point to V18. Did they have a similar problem while on V16.

Also is there anything running in the background that could cause the VM to temporarily "lock up"? Examples that we have heard of are e.g. when a Veeam backup is running, the VM "freezes" for a few seconds, or when a snapshot is being taken.