[Serusers] CDR Issues

Adrian Georgescu ag at ag-projects.com
Thu Dec 30 17:23:52 CET 2004

It is perfectly possible as the solution works as you described.


On Dec 30, 2004, at 5:16 PM, Nils Ohlmeier wrote:

> For the scenario we are talking here about, you can not guarantee 100%
> accurate accounting except you are using a B2BUA. Because if the call 
> is on
> hold it could stay in that state forever. If the UA's do not support 
> RTCP you
> will not see any traffic from them if they are restarted or cut off 
> from the
> net. So IMHO the only way would be to probe the UA's regularly with SIP
> requests which check if the dialog between those two parties is still 
> alive.
> And if you want to support all possible UA's this SIP request would 
> have to
> be an re-INVITE. If you then get an 408 or 481 from one partie you can 
> safely
> create the "call end" record. So at the end you need a B2BUA.
> Regards
>   Nils
> On Thursday 30 December 2004 16:05, Adrian Georgescu wrote:
>> Erik,
>> CDRTool guarantees 100% accurate accounting for call detail records
>> generated by SIP Express Router in combination  with MediaProxy
>> regardless of the presence of BYE messages. This solution is available
>> under commercial agreement.
>> http://www.ag-projects.com/CDRTool.html
>> Regards,
>> Adrian
>> OK, i've run into a problem with the CDR Creation.
>> If you have 2 users who call each other, talk a while and then instead
>> of hanging up decide to pull the plugs from their phones (or 
>> terminitate
>> their application), no complete CDR is generated, the INVITE and ACK 
>> are
>> logged, but since no one realy hangs up there won't be a BYE record, 
>> so
>> no CDR end/total time and no billable time.
>> Use the / a rtpproxy you would say, to bad it suffers from the same
>> problem, if both partes put the call on hold and pull the plug the 
>> call
>> keeps existing and once again an incomplete CDR.
>> Has anyone ever suffered the same problem? And what is the most 
>> reliable
>> way to generate CDRs?
>> Kind regards,
>> E. Versaevel
>> _______________________________________________
>> Serusers mailing list
>> serusers at lists.iptel.org
>> http://lists.iptel.org/mailman/listinfo/serusers

More information about the sr-users mailing list