Hello,
What if the values of the timers are absolute relative
to the first INVITE sent ?
Time
0 INVITE1 (retransmit after 500) (t0)
500 INVITE2 (retransmit after 1000) (relative to t0)
1000 INVITE3 (retransmit after 2000) (relative to t0)
2000 INVITE4 (retransmit after 4000) (relative to t0)
4000 INVITE5 (retransmit after 8000) (relative to t0)
8000 INVITE6 (retransmit after 16000) (relative to t0)
16000 INVITE7 (retransmit after 32000, but this is equal
to transaction timeout, so no retransmit) (relative to t0)
32000 Would send INVITE8, but transaction timed out.
In that case, the respective delays for response are 0.5,
0.5, 1, 2, 4, 8, and 16 seconds, with no odd delay at the end.
This seems to make more sense.
-Robin Lavall�e
Hi!
Please, correct me if I wrong.
draft-ietf-sip-rfc2543bis-09.txt
17.1.1.2
There are two timers A and B, first one for retransmissions and multiplies
each time by 2
starting from 500 ms and second one for transaction timeout and set to
64*500 and that amount of time is enough to retransmit request 7 times.
Let see what happens:
Time
0 INVITE1 (retransmit after 500)
500 INVITE2 (retransmit after 1000)
1500 INVITE3 (retransmit after 2000)
3500 INVITE4 (retransmit after 4000)
7500 INVITE5 (retransmit after 8000)
15500 INVI TE6 (retransmit after 16000)
31500 INVITE7 (do not retransmit)
32000 Transaction Timeout (Transaction terminated)
It means that INVITE7 have only 500 ms to be responded.
Is that correct?
If correct why after 1,2,4,8,16 seconds waiting we need one more half second
wait?
Sergey.
_______________________________________________
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
