Re: SIP Tracing Facility

Hadriel Kaplan <[email protected]>
Newsgroups gmane.ietf.sip
Message-ID <E6C2E8958BA59A4FB960963D475F7AC314C4C40D48@mail>

> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf Of Dale
> Worley
> Sent: Thursday, March 05, 2009 10:26 AM
>
> On Thu, 2009-03-05 at 08:16 +0000, Elwell, John wrote:
> > What about non-INVITEs? Provisional responses are designed for INVITEs
> > and I am not certain about their applicability to non-INVITEs.
>
> Yes, provisional responses to non-INVITEs may be slightly out of
> specification, but I don't expect any SIP element to malfunction if it
> receives a provisional response to a non-INVITE request.  OTOH, I would
> expect most uses of 'trace' to be attached to INVITEs, although that
> makes me think that tracing an OPTIONS might be a useful test tool.

OPTIONS doesn't always get routed through the same path as INVITE, fwiw.


> (Which leads to an unexpected observation:  RFC 3261 presupposes that
> all non-INVITE responses are processed promptly, but in the face of
> transport errors, this is not always so.  In sipX, we've adjusted
> transaction handling so that if a re-send of a non-INVITE message is
> seen the recipient sends a 100 to quench further re-sends.  It seems to
> work exactly as expected, and noticeably reduces re-sends in situations
> with transport errors.)

I agree for most methods, but we've seen problems with 100 for the OPTIONS.  Some devices use the OPTIONS as a reachability ping/test, so sometimes the 100 makes them think the far-end is reachable even if they get a 408 later; in other cases some devices look for specific response codes to detect administrative out-of-service, for example a 500/503 to an OPTIONS sometimes is used for that, and sometimes they didn't code for the 100 Trying as being a possibility, so they see it and assume it's not out-of-service, and the 503 sent later doesn't get processed the same.  I consider these bugs/broken implementations, but just noting sending the 100 is not bug-free.

 -hadriel
_______________________________________________
Sip mailing list  https://www.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use [email protected] for questions on current sip
Use [email protected] for new developments on the application of sip
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.