[sr-dev] nathelper generates a keepalive request with branch=0

[Klaus Darilion]

Maxim Sobolev wrote:
> Jan Janak wrote:
>> On Thu, Apr 29, 2010 at 6:30 AM, Maxim Sobolev <sobomax at sippysoft.com> 
>> wrote:
>>> Klaus Darilion wrote:
>>>>> Why not to use a SIP/2.0 compliant branch (z9hG4bK...)? any reason?
>>>> Maybe this is the method to detect keep-alive replies and absorb them
>>>> before they enter dialplan processing.
>>> I don't think it's intentional.
>>
>> There was a thread discussing this a long time ago, see:
>>
>> http://lists.iptel.org/pipermail/serusers/2005-April/018559.html
>>
>> Hope that helps.
> 
> It doesn't look like the same issue at all. The thread is about branch 
> in ACK being different from branch in 200 OK, while this one is about 
> branch in OPTIONS that the proxy generates to keep NAT binding alive.

Maybe it is related as the ACK is generated by the proxy itself and the 
OPTIONS too.

Anyway, I think branch=0 is not a clean solution, but RFC 3261 (17.2.3) 
defines that clients must support old transaction matching schema too.

regards
klaus