|
Hi,
Loop detection was made optional in RFC 3261. Now it
seems
that, due to a problem with its processing at
stateless proxies,
loop-detection cannot be performed
correctly (as it is given in RFC
3261) at stateless proxies.
I think loop detection is not a bad thing to have. Our
problem with
loop detection is not because loop detection itself has
problems.
It is because we tried to overuse the branch-id parameter
for too many things. The purpose
of branch-id parameter has
changed over time, but there seems to have been little or no
attempt to redesign
mechanism of loop-detection w.r.t. these
changes.
I suggest removing the loop detection part from the
branch-id
parameter and placing it in a new parameter. Having
loop detection
done using a separate parameter would avoid it
from interfering
with branch-id's primary use as transaction
identifier. All other
mechanism remaining the same as given in RFC 3261, I
think,
this will allow implementations to support loop
detection without
much trouble.
Thanks
Sachin
|