Re: [sipcore] 503 response with Retry-After header field - request for clarification

[Brett Tate <[e-mail]>]

> If you send a BYE to the same thing that sent you a 503
> for any other reason, before the Retry-After has expired,
> you are violating protocol.

Yes; it is a violation of the RFC 3261 section 21.5.4 SHOULD NOT.

"A client (proxy or UAC) receiving a 503 (Service Unavailable) SHOULD
attempt to forward the request to an alternate server.  It SHOULD NOT
forward any other requests to that server for the duration specified
in the Retry-After header field, if present."

RFC 5390 presents some reasons (directly or indirectly) why you might want
to violate the SHOULD NOT.  For instance, you might consider the 503 with
Retry-After mechanism to be too insecure to trust that you really should
not send any requests to a particular server for the next hour, day, or
year.

Similarly as mentioned within RFC 5390 section 5.2, RFC 3261 is unclear
concerning "server" in relation to the scope of the 503.  It is unclear
about "server" being 1) IP address where sent request, 2) source IP
address of received response, and/or 3) something else.  And based upon
draft-ietf-sipcore-dns-dual-stack's intention to cause trying both IPv4
and IPv6 addresses, I'd love to hear the proposal about how you avoid
attempting the same server twice when a 503 is received.

Thanks,
Brett