Bad Description Of Call Ref Value In H.323v2?

[Pete Cordell]

Tom,

CRV has a long and confusing history.  Originally the CRV on the Q.931
parts and the RAS parts were intended to be the same.  CRV along with
CID were expected to be unique.  We went for decision, and then it was
found that multiple parties in a conference could potentially assign the
same CRV.  This wasn't really a problem for the Q.931 channel but was
for a gatekeeper that was using this information to decide which record
it was looking at.  In Boulder srcCallSignalAddress was added into the
SETUP message as part of the implementor's guide (v1.1 if you like) so
that the combination of the CRV and srcCallSignalAddress could be used
to identify a call.  i.e. CRV was is a unique number generated by the
uniquely addressed entity defined by srcCallSignalAddress.  This worked
for ARQ but since srcCallSignalAddress didn't appear in BRQ or DRQ it
wasn't a complete solution.  Hence if a called endpoint receives a call
with a CRV that it is already using in a currently valid ARQ it has to
generate a new and unique number.  Hence CRV is different in the Q.931
and RAS channels.  To try and tie it all back together again the
CallIdentifier has been added in V2.

So the short answer is that CRV in Q.931 and RAS may/will be different
(assuming that was the question).

Moral of the story....  little slips cause big problems, and slips do
happen!!!  So if you've got any free time (whatever that is!) please
look at the draft and try to weed out any issues!!!

Regards,

Pete
=================================
Pete Cordell
BT Labs
E-Mail: pete.cordell at bt-sys.bt.co.uk
Tel: +44 1473 646436
Fax: +44 1473 645499
=================================


>----------
>From:  Tom Taylor[SMTP:Tom.Taylor.taylor at NT.COM]
>Sent:  11 December 1997 13:00
>To:    ITU-SG16 at MAILBAG.INTEL.COM
>Subject:       FW: Bad Description Of Call Ref Value In H.323v2?
>
>There is a problem with the wording of H.323v2, Section 7.4, when it
>talks about Call Reference Value.  From the lack of reaction, I suspect
>my earlier message did not go through.
>
>>-----Original Message-----
>>From:  Taylor, Tom (P.T.)
>>Sent:  Tuesday, November 11, 1997 3:33 PM
>>To:    'itu-sg16 at mailbag.intel.com'
>>Subject:       Bad Description Of Call Ref Value In H.323v2?
>>
>>Section 7.4 of H.323v2 talks about Call Reference Value as it appears in
>>Q.931 and RAS messages.  It indicates that the RAS channel itself has a CRV
>>-- not true, surely.  Perusal of H.225.0 makes clear that the CRVs in
>>the RAS
>>messages are copied from the Q.931 calls they are about, and are used to tie
>>the RAS messaging to the latter.
>