[OpenSER-Devel] Possible bug in the tm module in the presence of packet loss/branches

[Maxim Sobolev]

Dan Pascu wrote:
> My issue with what you're proposing is that it tries to modify the SIP 
> callflow to something no specified in the RFC, to solve a non-SIP 
> problem. I also do not like the idea that the proxy would keep 
> retransmitting on a branch after the originator has canceled the call.

Well, the problem is that while it's not specified directly in the RFC, 
but the current UAC behavior clearly violates 3-way handshake mechanism 
and leaves window open for UAC is UAS ending up in disjoint states due 
to 100 Trying loss. We can argue how big that window is (for example on 
some of my servers I can see about 0.1% of UDP being dropped), but the 
fact remains.

If you look at the Figure 5 in the RFC 3261 the only way client 
transaction could move directly from the Calling state to the Terminated 
state is via Timer B or Transport Err. No other path exist.

Regards,
-- 
Maksym Sobolyev
Sippy Software, Inc.
Internet Telephony (VoIP) Experts
T/F: +1-646-651-1110
Web: http://www.sippysoft.com