Now that I think about it a little more, your interpretation is right and the text in RFC3261 is actually buggy.
I think it should be interpreted the way I said; that the last (7th) transmission should occur 16 seconds after the first initial one, not 16 seconds after the 6th one. So something like this: --> INV 500 ms = T1 --> INV 500 ms (1 second from the initial transmission, 2*T1) --> INV 1 sec (2 second from the initial transmission, 4*T1) --> INV 2 sec (4 second from the initial transmission, 8*T1) --> INV 4 sec (8 second from the initial transmission, 16*T1) --> INV 8 sec (�6 second from the initial transmission, 32*T1) --> INV 16 sec (32 second from the initial transmission, 64*T1) Timer B fires. So after the first retransmission, timer A is reset to T1 again. From there on Timer A resets to double the previous value. /Hisham > -----Original Message----- > From: ext Christian Jansson [mailto:[EMAIL PROTECTED] > Sent: Wednesday, August 27, 2003 3:54 PM > To: Khartabil Hisham (NMP/Helsinki); [EMAIL PROTECTED]; > [EMAIL PROTECTED] > Subject: RE: [Sip-implementors] The seventh INVITE > transmission - isn't > itawaste > > > > > > The 7th transmission is 32*T1 (assuming T1 is 500ms, the 7th > > transmission occurs 16 seconds before the termination of > transaction). > > > > /Hisham > > > No. The 7th transmission is 32*T1 after the 6th transmission, so in > absolute time the 7th transmission is 63*T1 = 31,5s, that is > T1 (500ms) > before TimerB = 64*T1 fires. Look at the picture below. > > I also start to wonder why we wait 16s between the 5th and the 6th > retransmission but only waits 500ms after that before terminating the > transaction. > > > -------> INVITE sent for the first time at 0s > | > | T1 = 500ms > | > ---> 1st retransmission after T1 = 500ms > | > | 2*T1 = 1s > | > ---> 2nd retransmission after 3*T1 = 1,5s > | > | 4*T1 = 2s > | > ---> 3rd retransmission after 7*T1 = 3,5s > | > | 8*T1 = 4s > | > ---> 4th retransmission after 15*T1 = 7,5s > | > | 16*T1 = 8s > | > ---> 5th retransmission after 31*T1 = 15,5s > | > | 32*T1 = 16s > | > ---> 6th retransmission after 63*T1 = 31,5s > *<--- TimerB Fires at 64*T1 = 32s > > > > / Christian Jansson > > _______________________________________________ > Sip-implementors mailing list > [EMAIL PROTECTED] > http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors > _______________________________________________ Sip-implementors mailing list [EMAIL PROTECTED] http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors
