3CX IVR service stopped randomly on version 15.5 sp6

[iversion-Support]

For one of my deployment 3CX IVR service stopped randomly from last few weeks. The system was running fine for few months in AWS on server 2016. Plenty of CPU and RAM no issue there. I have already backup the system, uninstall and re-installed it but the issue is not resolved.

The last change made was by the client, he created few new IVR and call flow rules, I noticed for couple of IVR he sat the no input time to one second and action set to transfer the call to call queue.

The error i get in the windows event log is:


Faulting application name: 3CXIVR.exe, version: 0.0.0.0, time stamp: 0x5b838d74
Faulting module name: ntdll.dll, version: 10.0.14393.2608, time stamp: 0x5bd133d4
Exception code: 0xc0000374
Fault offset: 0x00000000000f7b43
Faulting process id: 0x8dc
Faulting application start time: 0x01d4f6b26983ddcb
Faulting application path: C:\Program Files\3CX Phone System\Instance1\Bin\3CXIVR.exe
Faulting module path: C:\Windows\SYSTEM32\ntdll.dll
Report Id: b4b4cfea-9bf7-4d56-98a3-2154fc3f2e0a
Faulting package full name:
Faulting package-relative application ID:

The 3CX system event around the time of failure is below:

[EC100009: Extenal application [computer name :0/3CXexhost] is disconnected]

My best guest is it has to do with short timer on IVR, My action plan is to update the system to version 16 and also add at least 3 seconds no input time for IVR and observe.

My workaround in the meantime is to set the recovery option for IVR set to restart after 1 minute of failure to minimize the downtime.

Anybody got any idea, have you seen similar issue.

Fahad

 

[iversion-Support]

Just' to update, I have change the minimum IVR timer to 3 seconds, but I also ran sfc /scannow on the server. It came back with alot of repairs entries in the log, including some of .net stuff. Given that its not a clean windows install but AWS image I figured the windows update or left over stuff in window may have corrupted some of the files.

The good new is sfc /scan now reported fixing all of the files 100% success. Will observe the system over the next few days to see if service crash again or not.

 

[StefanW]

FYi - we never tried 3CX on Windows on AWS. Only Debian 9...

 

[iversion-Support]

Thanks for letting me know, I don't think Windows Server or AWS is an issue, the same server ran for like two months without any troubles. I am confident it has to do with client setting up IVR with if no response 1 second transfer call. I think its just such a small timer for IVR specially running in cloud to cause service to fail when under load.

I also have another 3CX deployed in AWS on windows server 2016 with quarter the server resources no crash.

The difference is that one is v16. I will observer and let you know if the service fail again in next few days.