Switching to H.245

Sasha Ruditsky sasha at RADVISION.COM
Fri May 18 15:16:25 EDT 2001

Hi François

Probably I'll try to state you question in more generic way.

We know that it is possible to perform OLC operations after fast start
and before another Q.931 message (like ALERTING or CONNECT) with fast start
is sent.
(the simplest probably is to close channel opened using fast start)
(3rd party pause and redirection is of course procedure of this kind)

What is the meaning of the fast start in such CONNECT message?

And here I agree with Chris that such fast start elements should be ignored.

So probably the text that may solve this and should be added should state
something like:
"H.323 entity shall process fast start elements only once during the call."

BTW Probably the same statement should be done about h245Address. I just
know one H.323 endpoint that established 3 H.245 connections when got the


-----Original Message-----
From: Chris Wayman Purvis [mailto:cwp at ISDN-COMMS.CO.UK]
Sent: Friday, May 18, 2001 6:56 AM
To: ITU-SG16 at mailbag.cps.INTEL.COM
Subject: Re: Switching to H.245


Good catch!  My suggestion would be to add some text saying that anything
exchanged via H.245 overrides anything done by fastStart, even if the
fastStart happens to arrive later.

Now what happens if there's a possible scenario someone can devise where you
manage to get tunnelled and "separate connection" H.245 contradicting each
other is a thornier problem - can anyone see that happening?


Francois Audet wrote:

> Hum, interesting.
> My follow up question is then:
>     * A sends SETUP to B with fast start with h245Address and
>       tunnelling. ALERTING responds to FS, and the whole TCS and M/S
>       process takes place. Then A wants to forward the call (internally)
>       before answer (i.e., in the ALERTING phase). It thus sends TCS=0,
>       TCS=full, OLCs and all that. Then B answers and sends the CONNECT
>       (with the same fastStart as the ALERTING as per earlier
>       discussions we had ;^)  Wouldn't the content of the fastStart
>       contradict what the actual forwarded connections is really? Would
>       it be a problem?
> Seems pretty tricky...
>     -----Original Message-----
>     From: Paul Long [mailto:plong at IPDIALOG.COM]
>     Sent: Thursday, May 17, 2001 11:52 AM
>     Subject: Re: Switching to H.245
>     Francois,
>     I realize how this could be confusing, but I don't necessarily see a
>     conflict since they all say, "may." For example, if I say, "X may do
>     A or B," and, "X may do B," that does not preclude "X may do A."
>     Conversely, if I said "shall" instead of "may," I think there would
>     be real conflicts. I've worked on EPs that are very aggressive
>     during call establishment. For example, Setup contains an
>     h245Address and indicates support for Fast Connect and H.245
>     Tunneling. Other than all the non-compliant EPs out there, it worked
>     just fine. The EPs also support third-party pause, but I don't
>     remember ever testing that particular scenario. As long as an EP is
>     implemented correctly, I don't see anything in the Recommendation
>     that would prevent this from working.
>     Paul Long
>     ipDialog, Inc.
>         -----Original Message-----
>         From: Mailing list for parties associated with ITU-T Study Group
>         16 [mailto:ITU-SG16 at mailbag.cps.INTEL.COM]On Behalf Of Francois
>         Audet
>         Sent: Thursday, May 17, 2001 1:07 PM
>         To: ITU-SG16 at mailbag.cps.INTEL.COM
>         Subject: Switching to H.245
>         Guys,
>         H.323/ says :
>             After establishment of a call using the Fast Connect
>             procedure, either endpoint may determine that it is
>             necessary to invoke call features that require the use of
>             H.245 procedures. Either endpoint may initiate the use of
>             H.245 procedures at any point during the call, using
>             tunnelling as described in 8.2..1 (if h245Tunnelling remains
>             enabled) or a separate H.245 connection. The process for
>             switching to a separate H.245 connection is described in
>         8.2..3 says:
>             When H.245 encapsulation or Fast Connect is being used,
>             either endpoint may choose to switch to using the separate
>             H.245 connection at any time.
>         There seem to be some contradiction in there: is it "after
>         establishment" or "at any time"?
>         Do you have to wait for after CONNECT to establish a separate
>         H.245 channel or not?
>         The case I'm interested in would be to send SETUP with
>         fastStart, then receive ALERTING with fastStart. Then can either
>         end initiate H.245  before CONNECT? If so, what if third party
>         pause and redirection  is initiated before CONNECT?
>         ----
>         François AUDET, Nortel Networks
>         mailto:audet at nortelnetworks.com, tel:+1 408 495 3756

Dr Chris Purvis -- Development Manager
ISDN Communications Ltd, The Stable Block, Ronans, Chavey Down Road
Winkfield Row, Berkshire.  RG42 6LY  ENGLAND
Phone: +44 1344 899 007
Fax:   +44 1344 899 001

For help on this mail list, send "HELP ITU-SG16" in a message to
listserv at mailbag.intel.com

For help on this mail list, send "HELP ITU-SG16" in a message to
listserv at mailbag.intel.com

More information about the sg16-avd mailing list