[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Sipping] New I-D: A solution for the HERFP caused by forked SIPINVITE requests



Dale R. Worley wrote:
I've been told (and I agree) that it is impossible for a proxy to
determine accurately whether a request is looping or not, since a
request may legitimately loop through a proxy if it is "different", and
the proxy really can't judge what differences will be significant to the
destination UAS or not.  So best current practice is for a proxy to not
do loop detection at all, but rather leave it to the UAS.

YMMV. I agree that looping in the 3G/IMS world with the request being shuttled between various servers is problematic and best turned off. However, loop detection as envisioned in rfc3261 (and rfc2543), at the very least, stops a spurious request from chewing up further network and endpoint resources as it was routed multiple times to its destination.

But insofar as it is the job of the proxy to route requests, it
should be able to factor in those headers that aid in routing to
determine whether a request has looped or legitimately spiraled.

Now merged request is another matter.  If I correctly recall from
the mists of time, we argued during rfc2543-bis whether proxies
should detect merged requests or not and reached the conclusion
that they should not.

- vijay

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use sip-implementors at cs.columbia.edu for questions on current sip
Use sip at ietf.org for new developments of core SIP