Even though your second record says that it is Sip-Method = Bye, it
really is ACK. The provided sip_dictionary does not match what SER is
sending.
Sip_dictionary says:
VALUE Sip-Method Other 0
VALUE Sip-Method Invite 1
VALUE Sip-Method Cancel 2
VALUE Sip-Method Ack 3
VALUE Sip-Method Bye 4
But SER is actually using (msg_parser.h):
enum request_method { METHOD_UNDEF=0, METHOD_INVITE=1, METHOD_CANCEL=2,
METHOD_ACK=4,
METHOD_BYE=8, METHOD_OTHER=16 };
Notice that 4 in the dictionary is Bye, but in msg_parser.h it is Ack.
Also notice that you receive a 3rd RADIUS accounting message with
Sip-Method = 8. That is the real BYE. So change sip_dictionary to:
VALUE Sip-Method Undefined 0
VALUE Sip-Method Invite 1
VALUE Sip-Method Cancel 2
VALUE Sip-Method Ack 4
VALUE Sip-Method Bye 8
VALUE Sip-Method Other 16
If you want to eliminate the ACK radius records add this to ser.cfg:
modparam("acc", "report_ack", 0)
I have yet to see a situation where the ACK record is different from
than the INVITE, other than the Sip-method type. I'm puzzled as to why
they send it.
Hope that helps.
- Robert Liao
-----------------------------------------
Jan Janak writes:
Is Sip-Method set to ACK in one of the start records ?
Jan.
NO.
I have two consecutive starts with:
Sip-Method = Invite
and
Sip-Method = Bye
and one stop with
Sip-Method = 8
Razvan Radu
On 06-10 16:31, SER Users wrote:
> hello,
>
> I have ser configured with radius accounting and for each call
> I get two Start packets and one Stop packet.
>
> Is this the correct behaviour or I have miss-configured something ?
> The only difference between the two Start Packets is the "Sip-Method"
> attribute which in one is "Invite" and in the other one is "Bye"
>
> In my configuration I have:
>
> ...
> modparam("acc", "radius_flag", 2)
> modparam("acc", "radius_missed_flag", 3)
> ...
> setflag(2);
> if (!t_relay()) {
> sl_reply_error();
> };
> ...
>
>
> thanks,
> Razvan Radu
>
>
> _______________________________________________
> Serusers mailing list
> Serusers at
iptel.org
>
http://lists.iptel.org/mailman/listinfo/serusers