Dear All,
I support the principle that only one H.225 call signalling message should be sent per TPKT. Firstly I believe it was agreed between early implementers that this is the way they would work. Secondly, even though it is possible to send multiple PER messages per TPKT there are other reasons why this is not totally desirable. In our case of working on a high volume GK, we only need to process a fraction of H.245 messages. If there is only one message per TPKT we can easily work out what the message is by looking at the first few bytes. However, if there are multiple messages per packet we have no choice but to decode each and every message. In most cases this represents a totally unnecessary processing overhead and is therefore best avoided.
Pete
============================================= Pete Cordell pete@tech-know-ware.com =============================================
-----Original Message----- From: Paul Long Plong@SMITHMICRO.COM To: 'h323implementors@imtc.org' h323implementors@imtc.org Cc: ITU-SG16@mailbag.cps.intel.com ITU-SG16@mailbag.cps.intel.com Date: 30 June 1999 17:53 Subject: RE: PDU & ASN parser
Guntram,
I raised this issue over a year ago. In addition, H.225.0 does not address this issue regarding RAS (it does regarding call signaling), so I suppose
one
could pack multiple RAS messages per unreliable-transport PDU. As you say,
it
doesn't make sense to pack messages, and I don't think anyone does it. We should probably add something like the following to section 6.1 of the next revision of H.225.0.
"When messages are sent on the H.225.0 RAS channel, only one whole message shall be sent within the boundaries defined by the transport; there shall
be
no fragmentation of H.225.0 RAS messages across transport PDUs."
We should also insert the indicated text, below, into existing text in the same section.
"When messages are sent on the reliable H.245 control channel, >>>exactly
one
message should be sent but<<< more than one message may be sent within the boundaries defined by the reliable transport PDU as long as whole messages
are
sent;"
Paul Long Smith Micro Software, Inc.
-----Original Message----- From: Guntram Diehl [SMTP:gdiehl@innovaphone.com] Sent: Wednesday, June 30, 1999 4:14 AM To: 'Bancroft Scott'; 'Srikanth Nayani' Cc: 'h323implementors@imtc.org' Subject: RE: PDU & ASN parser
Yes this can be done, but does any implementation actually do it? Why would anyone want to do this, it just saves 4 bytes tpkt header per additional H.245 message and it does not change the fragmentation of the whole thing by IP.
Or is there another reason why this was specified in the standard this way?
Guntram
-----Original Message----- From: Bancroft Scott [mailto:baos@OSS.COM] Sent: Wednesday, June 30, 1999 3:54 AM To: Srikanth Nayani Cc: h323implementors@imtc.org Subject: Re: PDU & ASN parser
On Tue, 29 Jun 1999, Srikanth Nayani wrote:
In the H.225 document chapter 6 mentions that multiple messages
can be
sent over the reliable H.245 control channel within the boundaries
of the
reliable transport channel . Does that mean that two H.245 PDUs
encoded
using PER and whose total encoded length is within the boundary of
the
H.245 TCP channel can be merely concatenated and sent as a single
payload
of the H.245 TCP channel. If so how does a ASN.1 decoder know that there are two
messages in the
payload and the offset at which the second message starts.
An PER decoder does not know that there are two messages in the payload. It does know where the first message ends, so the location of the next message can be derived from this knowledge.
-- Bancroft Scott Toll Free :1-888-OSS-ASN1 Open Systems Solutions, Inc. International:1-609-987-9073 baos@oss.com Tech Support :1-732-249-5107 http://www.oss.com Fax :1-732-249-4636
for list subscription instructions, send email to h323implementors-request@imtc.org with text "info"
for list subscription instructions, send email to h323implementors-request@imtc.org with text "info"
for list subscription instructions, send email to h323implementors-request@imtc.org with text "info"