The capture that you sent, just looks like a normal callflow. The only thing that stands out is the useless multiple 18x with increasing rseq. If this is typical, you might want to ask the vendor why they are sending it. For instance, the timer to send periodic 18x to refresh Timer C might be configured too low.
From: NK [mailto:[email protected]] Sent: Tuesday, April 09, 2013 7:22 AM To: Brett Tate Cc: [email protected] Subject: Re: [Sip-implementors] PRACK Query Hi Brett, Thanks for your help on this. I checked the TO and FROM Tag for both the transaction and they are matching, however if i checked the Branch ID, they are not matching on every PRACK coming from Ingress side. I have attached the traces for for 1 call, so that it can clearly describe the issue. Very big thanks for your help again. Regards, Nitin On Tue, Apr 9, 2013 at 5:52 AM, Brett Tate <[email protected]<mailto:[email protected]>> wrote: Hi, You haven't provided enough information for someone to indicate the issue. Are the To tags different? If so, does the PRACK match the correct dialog and did the PRACK's branch correctly change? Within a dialog, how many rseq values are there (i.e. how many new 18x responses instead of retries)? RFC 3262 indicates how the UAC and UAS work. It also indicates the retransmission rules. A retry of the same 18x (i.e. same To tag and rseq) should not cause a resend of the PRACK. Thus if the 18x is still being retried, it might not like the PRACK. A retry of the PRACK stops upon receiving the PRACK's final response. Thus if the PRACK is still being retried, it might not like the 200 response. Although potentially not the best way to deal with the atypical situation, this could be a reason for CANCEL based upon PRACK timeout instead of CANCEL because of other reason. > -----Original Message----- > From: > [email protected]<mailto:[email protected]> > [mailto:sip-<mailto:sip-> > [email protected]<mailto:[email protected]>] > On Behalf Of NK > Sent: Monday, April 08, 2013 5:39 PM > To: [email protected]<mailto:[email protected]> > Subject: [Sip-implementors] PRACK Query > > Dear All, > > I have query regarding the PRACK. > > I have the call scenario where my SBC send the INVITE to vendor > including > 100 rel and from vendor we received the rseq. Now I can see that my > customers sends 200 OK for PRACK. But after that continuously my vendor > is > sending 183 w/SDP or 180 w/SDP and asking PRACK and because of this > for a > long time PRACK exchange is happening and after few seconds my CUSTOMER > is > sending CANCEL to terminate the call. > > Is that the valid scenario? For how long vendor can continuous send > PRACK > request? I feel because of this I am getting LOW ACD aswell? Can you > please > advice? > > Thanks, > Nitin _______________________________________________ Sip-implementors mailing list [email protected] https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
