Re: M2PA: resynchronization of sequence numbers after PO

Dario,

Sorry, I was wrong.  Each side should respond LSR regardless of the state of
the local processor; however, because the local processor is out on the last
side recovering, it cannot necessarily synchronize sequence numbers (it takes
a flush from MTP3 to do that).

See Test Case 3.4.2 "RPO during LPO" and Test Case 3.4.3 "Clear LPO when "Both
processor outage" in draft-bidulock-sigtran-m2pa-test-08.txt

(It is no longer on the IETF servers, darn it expired again, but you can get a
copy here:

  http://www.openss7.org/docs/draft-bidulock-sigtran-m2pa-test-08.txt

)

--brian


Dario wrote:                                  (Wed, 21 Nov 2007 03:59:57)
> 
> Hi Brian,
> 
> so if I understood correctly, the resinchronization takes place only when
> there is no PO on either
> side.
> it would be good to clarify that in RFC since statement in RFC4165, section
> 4.1.4 has not been clear to me:
> " Upon receiving the Link Status Processor Recovered message, the M2PA in
> RPO SHALL respond
> with a Link Status Ready message on the User Data stream."
> 
> This is then coniditionaly true, only if there is no local PO on that M2PA
> side.
> 
> Regarding second question, how many LS Ready messages are then exchanged?
> 
> A side sends LS PR.
> B side sends LS PR.
> It seems to me that each side sending one LS Ready (stream 1) is sufficient
> to resynchronize the sequence numbers.
> Or is it 4 LS Ready messages (each side sending two LSR messages?).
> 
> thanks!
> Best regards, Dario
> 
> 
> 
> Brian F. G. Bidulock wrote:
> > 
> > Dario,
> > 
> > Dario wrote:                                 (Tue, 20 Nov 2007 07:32:34)
> >> 
> >> Hi,
> >> two more questions:
> >> 1. If both sides are in PO and one processor recovers and sends LS PR.
> >> Should the other side respond with 
> >> LS Ready (stream 1) although it is locally in PO?
> > 
> > No.  Ala Q.703 it sends LS PR when it recovers and only then both
> > sides exchange LSR.  Neither side resynchronizes until the link
> > becomes unblocks which requires no processor outage (local or remote).
> > 
> >> 2. again, both sides are in PO and both sides recover - each side sends
> >> LS
> >> PR. Should both sides exchange just one pair of LS Ready (stream 1)
> >> messages
> >> or should they perform resynchronization sequence independently?
> > 
> > Each side resynchronizes independently per Q.703.  The sequence when the
> > last terminal recovers is the same as though only one processor was out.
> > 
> > --brian
> > 
> > -- 
> > Brian F. G. Bidulock
> > [email protected]
> > http://www.openss7.org/
> > 
> > 
> > _______________________________________________
> > Sigtran mailing list
> > [email protected]
> > https://www1.ietf.org/mailman/listinfo/sigtran
> > 
> > 
> 
> -- 
> View this message in context: http://www.nabble.com/M2PA%3A-resynchronization-of-sequence-numbers-after-PO-tf4812071.html#a13875827
> Sent from the IETF - Sigtran mailing list archive at Nabble.com.
> 
> 
> 
> _______________________________________________
> Sigtran mailing list
> [email protected]
> https://www1.ietf.org/mailman/listinfo/sigtran

-- 
Brian F. G. Bidulock
[email protected]
http://www.openss7.org/