[ previous ] [ next ] [ threads ]
 To :  Allan Sandfeld Jensen <linux@c...>
 From :  Milan Rukavina <milan@h...>
 Subject :  Re: [yate] CDR Status 'accepted'?
 Date :  Mon, 07 Jun 2010 09:54:40 +0200
Seems like I solved the problem. In the call.answered message I had:
'cdrcreate'=>'false','cdrtrack'=>'false'

but when set
'cdrcreate'=>'no' without cdrtrack I get billtime. BTW. this parameter 
seems to be needed in order not to have invalid cdrs for ivr channel.

One more strange thing I noticed: billtime is almost always greater than 
duration for about 0.5 - 1s. Not sure why is that?

Thanks,
Milan
> Hi Allan,
> I'm not worried about status - I understand that part, but billtime is
> confusing me since it's = 0. Finished calls in status answer have
> billtime 0, and duration is almost equal to ring time - but I'm sure my
> php ivr application answered and executed correctly - so I suppose
> billtime should be > 0. On th other hand, when call is finished in
> status accepted it has regular billtime.
>
> I'm not sure if my ivr can do anything wrong to break cdr creation but
> seems like other php ivrs have regular bill time.
>
> I'm using postgres db for cdr and this is defined in register.conf
>
> Thanks,
> Milan
>   
>> No. As I said, you can not rely on the status in call.cdr messages. The status 
>> is updated by the latest status cdrbuilder have seen, but messages are not 
>> necessarily arriving in the correct order, so cdrbuilder may see call.ringing 
>> after it sees call.answered. This only affects the status parameter though, 
>> billtime, duration and ringtime parameter can all handle messages arriving out 
>> of order, so rely on them instead. 
>>
>> Duration > 0.000 means the call have been accepted.
>> Ringtime > 0.000 means the call have been ringing at one point.
>> Billtime > 0.000 means the call have been answered.
>> If you really need it. I have a patch that also orders the status, I'll see if 
>> I can extract it and post it. We need it to get correct status in dialog 
>> NOTIFY.
>>
>> Mvh
>> `Allan
>>
>>   
>>     
>
>