Big problems with out.-calls after switching from 3CX12.5 (windows server) to 3CX15.5 (linux server)

Discussion in '3CX Phone System - General' started by ziead, Aug 18, 2017.

Thread Status:
Not open for further replies.
  1. ziead

    Joined:
    Aug 9, 2017
    Messages:
    4
    Likes Received:
    0
    Hello,

    we've upgraded from 3CX Phone System 12.5 Professional to current 15.5 Professional, but we're experiencing big problems that we're not able to solve (trying since monday).


    Main problem (affects outbound/outgoing calls):
    Often when a user tries to call an external phone number (one of our customers for example) -> the other/external phone rings(!) but the user in our company doesn't hear anything, no dial sound, no sound while the phone is ringing... and when this happens -> the called number dissapears in the 3CX softphone (in fact the user is somehow connected to the external phone/customer because the other phone is ringing and the client also looks like you are connected (red "hang up" button is shown...), except that the called number is not visible in the client (it dissapears when starting the call, not when entering the number).

    This does NOT happen all the time (sometimes eyerything works), but it happens often and strangely not to all users (for now) and for some more often than for others... (about 8 users had this issue about 1 - 3 times this week, 2 users have this 20 - 30 times day (or more)!). It does not seem related to the PC, the user account or the (physical) network connection (tested all of this, switched computers and user accounts and so on...).

    If this happens, the user have to try it 3 - 5 times, than it works (in most cases).


    Another problem (perhaps related to the problem above):
    Sometimes (not so often!) if a user tries to call an external phone/customer, he's just getting a "call could not be established" (or perhaps "could not establish call", we use the german version here)... If this happens, the user have to try it multiple times, than it works (after a while).


    Perhaps an important fact:
    Especially for the two clients which heavily suffer from the "main problem" above, we see a lot of 3CX server events like "Extension 24 is unregistered" & Extension 24 is registered", again and again, all the time... (but the network connection itself was already checked and is stable).


    Some information about our setup:
    Like I said we use 3CX Phone System 15.5 Professional now, with the 3CX Windows clients (softphones). Version 12.5 (on a Windows Server 2008R2 VM) in combination with older Windows 3CX clients worked without such problems (we've decided to upgrade because of security updates by 3CX).
    The server is a virtual machine with "Debian 8" Linux (like suggested by 3CX) on Microsoft Hyper-V (Windows Server 2012R2 host).
    We've installed 3CX Phone System (and imported our configuration backup from 12.5 and the former Windows Server) as it is explained on the official 3CX website. Of course we also configured the firewall as described by 3CX (but have also already desabled the firewall completely afterwards, just for testing/because of the described problems).



    Any help or hint would be appreciated!

    Thank you very much for reading!



    Best Regards
     
  2. eddv123

    eddv123 Active Member

    Joined:
    Aug 15, 2017
    Messages:
    916
    Likes Received:
    140
    Hi Ziead,

    If this issue only effects the soft client it maybe worth checking you are running the latest compatible version of client for your version of 3CX, as well as checking the device you are running it on has the latest updates also (if Windows Client then check the machine for Windows updates & .NET Framework).

    Is there any local firewall or Anti-virus software running on these PC's ? you could test with these turned off. Also check if there is any other applications running in the background which might clash with the port numbers being used by 3CX/3CX Client.

    Are you running the soft clients on their own or in CTI mode with a desktop phone ? What OS are they running on also ?

    Maybe worth checking that your VoIP Provider template is still supported (If using SIP Trunks) as I know there have been changes to supported VoIP providers and settings required in the templates recently. However it sounds like these calls are not reaching this far in these cases.

    I think the best thing you could do is try and capture the issue with your VERBOSE logging turned on and a PCAP trace - it might be able to give you an indication of what is going on here.
     
  3. ziead

    Joined:
    Aug 9, 2017
    Messages:
    4
    Likes Received:
    0
    Hi eddv123,

    thank you very much for reply & your suggestions!

    About your questions:
    We're using primary the 3CX Windows client/softphone alone, only some users have a CTI/desktop phone (the users who suffer the most/are primary affected by the described problems only use the 3CX client/softphone).

    Nearly all our users still use Windows 7 Professional x64, Security-Updates for Windows are automatically installed but optional .NET-Updates and such stuff may not be all installed at the moment - I will check this.

    On the client computers we use 'ESET Endpoint Security' for years now, it would be strange if it causes the problem (because we all use it here - nothing different on the two most problematic clients I mentioned) - But I will test it.

    Other applications are not really used on most of the clients (because they're working on a terminalserver anyway - 3CX client and a RDP terminalserver session are the main or only used "applications" most of the time).

    We don't use SIP trunks, we use a gateway (a "beroFix" hardware device) which is directly connected to hardware provided by our telephone company & ISP ("Deutsche Telekom"), everything located in our local server room and connected to the "Deutsche Telekom" via Fiber...

    I will have a look at verbose logging and trying a pcap trace...

    Interesting thing I've noticed after my first post:
    If everything works, in the active calls you see something like "10005 berofix E1 ("...number of the customer...")", the text before the phone number in brackets is the name of the gateway.
    BUT when the error occurs (the 'Main problem') -> the destination column (not sure how it's name in the english version) ONLY shows the phone number (without brackets), no gateway name...

    Thanks & Best Regards!
     
  4. eddv123

    eddv123 Active Member

    Joined:
    Aug 15, 2017
    Messages:
    916
    Likes Received:
    140
    Additionally If you are using a beroNet gateway then I would also run the dialplan debug tool in the gateway to see if when this issue presents itself, whether the call actually reaches the gateway - it will tell you quite clearly in the the interface.

    It would give a good indication of where the problem might lie.
     
  5. ziead

    Joined:
    Aug 9, 2017
    Messages:
    4
    Likes Received:
    0
    Hi eddv123,

    the debug tool in the beroNet webinterface doesn't work, it just won't start debugging (instead just reloads the page) - don't know why...

    But anyway: Thanks for your help!
     
  6. eddv123

    eddv123 Active Member

    Joined:
    Aug 15, 2017
    Messages:
    916
    Likes Received:
    140
    Hi Ziead,

    Firstly check you are either running in Google Incognito mode or have cleared you Web Cache and then login to the beroNet device.

    Go to Management >> Dialplan Debug >> And start the Dialplan Debug.

    This should pop-up into a new window unless you have pop-ups blocked from your web browser - you need to disable pop-up blocker if this is the case.

    This new window should have State: ON which tells you that you can replicated your issue.

    Replicate and in real time it will give you an output (if a call is reaching as far as the gateway obviously). If you see nothing you know that the call problem does not lie with the Gateway in this instance.

    Let me know how you get on.
     
  7. ziead

    Joined:
    Aug 9, 2017
    Messages:
    4
    Likes Received:
    0
    Hi eddv123,

    thanks for reply! For some reason debugging worked only with IE here (not with other browsers, no matter their settings)...

    Interesting - When a call fails the used channel in the beronet device is "0" (always):
    CDR,1491,SIP:192.168.110.11,ISDN:5:0,"XXXXXXX" ,"" ,13,018041XXXXX,17/08/25-14:37:03,17/08/25-14:37:14,-,-,SIP,NUA_I_CANCEL:200,-,-
    S CANCEL|INDICATION: from="XXXXXXX" 13@192.168.110.11, to= "" 018041XXXXX@192.168.110.231
    I SETUP|REQUEST: port=5, channel=0, dad=018041XXXXX, oad=13
    D INCOMING src:13 dest:018041XXXXX -- OUTGOING src:13 dest:018041XXXXX
    S INVITE|INDICATION: from="XXXXXXX" 13@192.168.110.11, to=018041XXXXX@192.168.110.231


    When everything works/the call goes throug a different channel is used:
    CDR,1490,SIP:192.168.110.11,ISDN:5:3,"XXXXXXX" ,"" ,13,018041XXXXX,17/08/25-14:36:39,17/08/25-14:37:01,17/08/25-14:36:52,17/08/25-14:37:01,SIP,NUA_I_BYE:200,AOCD_U:4,-
    I RELEASE|INDICATION: port=5, channel=3, dad=018041XXXXX, oad=13, cause=16 (Normal call clearing)
    I DISCONNECT|REQUEST: port=5, channel=3, dad=018041XXXXX, oad=13, out_cause=16 (Normal call clearing)
    S BYE|INDICATION: from="XXXXXXX" 13@192.168.110.11, to= "" 018041XXXXX@192.168.110.231
    I CONNECT|INDICATION: port=5, channel=3, dad=018041XXXXX, oad=13


    I've replaced names and parts of numbers with X, but both calls were identical (same user, same external number/target).

    The reason seems to be a 'call' (?) showing up in the active channel list of the beronet device for channel '0' with empty 'DAD' and 'OAD' fields and call type 'g711_20'. This entry was permanently in the list (even while nobody was using the 3CX phone), if a call worked it was also in the list, fields were then filled and call type was 'voice'...
    The entry/call dissapeared from the beronet device when I rebooted it, and the problem was solved... but for how long?
    I've rebooted the berofix device (just to be sure you know) already multiple times the last days... and sometimes I thought the problem was gone (after several changes to the 3CX server) but it came back - So I think this strange 'ghost call' on the berofix - if it was/is really the reason - will come back and cause the whole problem again... (I don't really know why this problem started after we've upgraded the 3CX - Perhaps just a coincidence? Or some kind of incompatibility between 3CX v15 and the bero*fix/beronet 400 VoIP gateway...? Strange...)

    If you have any idea about this (or similar experiences), please let me know.

    Best Regards & Thank you very much for your time and your efforts!
     
Thread Status:
Not open for further replies.