RE: M2PA: Link alignment when proving omitted
"Soni Goel" <[email protected]>
| Newsgroups | gmane.ietf.sigtran |
|---|---|
| Message-ID | <5F76BCC269A4BC44AD6200842CBD531201737D14@sd-exchange> |
Hi Brian, Your question: OTOH, why is M2PA-1 sending LSA "in response to Out of Service"? LSA is sent in response to "Start". Make M2PA-1 ignore LSOOS in Not-Aligned state. Answer: M2PA-2 has not yet received start. M2PA-1 needs to restart Timer T2 and resend LSA, otherwise T2 will expire at M2PA-1 causing link failure. Later when M2PA-2 receives start and sends LSA, M2PA-1 will respond with only LSOOS. This is also as per following test case in draft-bidulock-sigtran-m2pa-test-08.txt - 3.1.2. Timer T2 ___________________________________________________________________ | | | CPT: IOT: SP B SP A | | VAT: PT IUT | | | | <----------------- OUT-OF-SERVICE | | OUT-OF-SERVICE -----------------> | | :start | | <----------------- ALIGNMENT | | (Note) OUT-OF-SERVICE -----------------> | | <----------------- ALIGNMENT (Note) | | ! | | ! T2 5.0 <= T2 <= 150.0 | | ! | | <----------------- OUT-OF-SERVICE | | !out of service(AERM) | | | |___________________________________________________________________| Thanks, Soni -----Original Message----- From: Brian F. G. Bidulock [mailto:[email protected]] Sent: Friday, June 29, 2007 12:02 PM To: Soni Goel Cc: [email protected] Subject: Re: [Sigtran] M2PA: Link alignment when proving omitted Soni, Soni Goel wrote: (Fri, 29 Jun 2007 10:57:24) > Hi, > > This is regarding the Link alignment when proving is omitted. > As per message flow given below, link is not able to come in-service > because M2PA-2 receives Link status Align after sending Link Status > Ready. > > I would suggest M2PA-2 to resend Ready on receiving Align, while timer > T1 is running. This is as per following statement in RFC 4165, Section > 4.1.3 Link Alignment - > "If the Link Status Ready message is sent, then M2PA MAY send additional > Link Status Ready messages while timer T1 is running." > > Please let me know your feedback. > > > MTP3-1 M2PA-1 SCTP-1 SCTP-2 M2PA-2 MTP3-2 > ---- ---- ---- ---- ---- ---- > . . . . . . > Start . . . . . > ------------> . . . . > . Associate . . . . > . ------------> . . . > . . . . . . > . . (SCTP Association . . > . . procedure) . . > . . . . . . > . Communication Up Communication Up . > . <------------ ------------> . > . . . . . . > . Link Status Out of Service . . > . ------------------------------------> . > . . . . . . > . Link Status Alignment . . . > . ------------------------------------> . > . . . . . . > . Start timer T2 . . . > . . . . . . > . . Link Status Out of Service . > . <------------------------------------ . > . Link Status Alignment . . . > . ------------------------------------> . > . . Link Status Out of Service . > . <------------------------------------ . > . Link Status Alignment . . . > . ------------------------------------> . > . . Link Status Out of Service(2) . > . <------------------------------------ . > . . . . . Start > . . . . <------------ > . . . Link Status Alignment(2) . > . <------------------------------------ . > . . . . . . > . Link Status Alignment(response to Out of Service(2)) . > . ------------------------------------> . OTOH, why is M2PA-1 sending LSA "in response to Out of Service"? LSA is sent in response to "Start". Make M2PA-1 ignore LSOOS in Not-Aligned state. --brian > . . . . Start timer T1 > . . . Link Status Ready(2) . > . <------------------------------------ . > . . . . . . > . Stop timer T2 . . . > . Link Status Alignment(response to Alignment(2)) . > . ------------------------------------> . > . . . . Link fails due to . > . . . . Alignment received in. > . . . . Ready state . > . . . . . > . . Link Status Out of Service . > . <------------------------------------ . > . . . . . . > . . . . . . > . Link inservice(received Ready(2)) . . > . Link Status Ready(response to Ready(2)) . . > . ------------------------------------> . > . . . . . . > . . . . . . > . Link Failure (received Out of Service from M2PA-2) . > . Link Status Out of Service . . > . ------------------------------------> . > . . . . . . > . . . . . . > > Thanks, > Soni > > _______________________________________________ > Sigtran mailing list > [email protected] > https://www1.ietf.org/mailman/listinfo/sigtran -- Brian F. G. Bidulock [email protected] http://www.openss7.org/