[asterisk-dev] SIP Timers and "100 Trying" on a Re-INVITE. Resource leak.

[Steve Davies]

On 13 June 2012 16:32, Kevin P. Fleming <kpfleming at digium.com> wrote:
> On 06/13/2012 10:27 AM, Steve Davies wrote:
>
>> Should I raise this as a JIRA ticket (with the cleaned-up grammar)?
>
>
> Yes, I think you should, and the summary can be as simple as "SIP re-INVITEs
> have no transaction timeout" or similar. I've already notified the rest of
> the Asterisk development team here at Digium about it, so they'll be
> watching for the issue to be entered.
>
> As best I can tell, this is the *only* SIP request method where this problem
> can happen, because it is the only one that allows provisional responses.

I agree. It can be caused by any (and all) of the provisional
responses, all of which cancel the destroy timer and assume that a
further response packet will arrive, but the only place this happens
in quite this way appears to on an INVITE.

Strictly speaking, the INVITE and the Re-INVITE suffer from the same
problem, but an initial INVITE is usually under the control of an
application such as Dial() or Queue() which has its own overriding
time-outs built in to save-the-day. The Re-INVITE is more spontaneous,
so has nothing to rescue it from leaking the resource.

  https://issues.asterisk.org/jira/browse/ASTERISK-19992

Regards,
Steve