Re: Re:Re: M3UA : Unexpected ASP-DOWNACK Message (Brian F. G. Bidulock)
"Brian F. G. Bidulock" <[email protected]>
| Newsgroups | gmane.ietf.sigtran |
|---|---|
| Organization | http://www.openss7.org/ |
| Message-ID | <[email protected]> |
chen.puran, [email protected] wrote: (Wed, 20 Jun 2007 15:52:41) > thanks brain; > > questionS: > > 1: > > question one: > If the ASP receives an unexpected ASP INACTIVE Ack message, how the ASP > should to do? > > the ASP should consider itself in the ASP-INACTIVE state? Read the RFC. RFC 4666 Section 4.3.4.4 ASP Inactive Procedures ... At the ASP, the ASP Inactive Ack message received is not acknowledged. Layer Management is informed with an M-ASP_INACTIVE confirm primitive. If the ASP receives an ASP Inactive Ack without having sent an ASP Inactive message, the ASP should now consider itself to be in the ASP-INACTIVE state. If the ASP was previously in the ASP-ACTIVE state, the ASP should then initiate procedures to return itself to its previous state. ... > > 2: > In RFC4666 SECTION 4.3.4.1. ASP Up Procedures > > <<If the ASP receives an unexpected ASP Up Ack message, the ASP should > consider itself in the ASP-INACTIVE state. If the ASP was not in the > ASP-INACTIVE state, it SHOULD send an Error message and then initiate > procedures to return itself to its previous state.>> > > question Two: > if ASP in in the ASP-DOWN status,it can consider itself in the ASP-INACTIVE > state? Yes, and then "initiate procedures to return itself to its previous state": that is, send an ASP-DOWN message. > i think how to deal wiht the unexpented ASPSM/ASPTM Message ,rfc466 is not > maturity. Did you have a proposal for improvement? --brian > > > ===================================================== > Chen, > > RFC 4666 Section 4.3.4.2 ASP-Down Procedure: > > ... > > At the ASP, the ASP Down Ack message received is not acknowledged. > Layer Management is informed with an M-ASP_DOWN confirm primitive. > If the ASP receives an ASP Down Ack without having sent an ASP Down > message, the ASP should now consider itself to be in the ASP-DOWN > state. > > If the ASP was previously in the ASP-ACTIVE or ASP-INACTIVE state, > the ASP should then initiate procedures to return itself to its > previous state. > > ... > > ...which would pretty much consist of sending ASP Up. > > This behavior was specifically design to permit what we have referred to > as the Unsolicited ASP Down Ack. The SGP may wish to autonomously send > an Unsolicited ASP Down Ack in the event that it is experiencing an > extensive failure (e.g. local isolation from the NIF) that makes it > impossible for it to serve any AS. This will drive all AS inactive for > the SGP and cause the ASP to reinitiate the ASP Up procedure. If the > isolation persists, the SGP may respond to the ASP Up procedure with an > ERR("Refused - Management Blocking") response per section 4.3.4.1 ASP Up > Procedures (below) until the local fault clears. > > ... > > If for any local reason (e.g., management lockout) the SGP cannot > respond with an ASP Up Ack message, the SGP responds to an ASP Up > message with an Error message with the reason "Refused - Management > Blocking". > > ... > > > Of course, an alternative would be to drop the SCTP association (but > this could result in significantly more message loss.) > > --brian > > > _______________________________________________ > Sigtran mailing list > [email protected] > https://www1.ietf.org/mailman/listinfo/sigtran -- Brian F. G. Bidulock [email protected] http://www.openss7.org/