Re: [Megaco] Modify to do Audit hook state

[Kevin Boyle <kboyle at huawei.com>]

Hi Raphael,

I concur with your assessment.  Honestly, it is difficult for me to
reconstruct my thinking when I made this response 5 years ago (!), though I
imagine that my thinking was along the lines you have expounded here, and
for some reason I was focusing only on the first case.

I don't really work on H.248 anymore, though, so I am a bit out of touch
with recent developments. :(

Kevin 

-----Original Message-----
From: megaco-bounces at ietf.org [mailto:megaco-bounces at ietf.org] On Behalf Of
Raphael Tryster
Sent: Sunday, August 30, 2009 8:45 AM
To: Kevin Boyle; Pascal Lambers; megaco at ietf.org
Subject: Re: [Megaco] Modify to do Audit hook state

Hi Kevin,

I would like to clarify the "yes" answer you gave to Pascal's question 5
years ago, because I think the answer to his precise question is "no".

What I mean is: "Yes", if you are an MGC and want to use Modify to audit the
hook state, then you have to include any events previously present in the
Events descriptor in order to guarantee their continued detection.

However: "No", it is NOT true that if the MG returns an error "termination
already off/on hook", the MGC has no longer any guarantee as to which events
are still programmed on the termination.  This is because IN THE SPECIFIC
CASE THAT THE MG HAS REPLIED WITH ERROR 540, the MGC KNOWS that the old
events descriptor was NOT replaced, because the command failed.

The motivation for this clarification is to verify that an MG that returns
error 540 and later notifies an event based on the events descriptor in the
command that failed, is behaving incorrectly.

Regards,
Raphael

-----Original Message-----
From: Kevin Boyle [mailto:kboyle at nortelnetworks.com]
Sent: Wednesday, 18 August 2004 12:50 AM
To: Pascal Lambers; megaco at ietf.org
Subject: RE: [Megaco] Modify to do Audit hook state

The MGC would have to include any events previously present in the Events
descriptor in order to guarantee their continued detection, yes.

Kevin 

-----Original Message-----
From: Pascal Lambers [mailto:pascal.lambers at pandora.be]
Sent: Tuesday, August 17, 2004 5:03 PM
To: megaco at ietf.org
Subject: Re: [Megaco] Modify to do Audit hook state

Kevin,

"The MGC must account for these kinds of issues when it is building
messages".
By this you mean, that if the MG returns an error "termination allready
off/on hook", the MGC has no longer any guarantee as to which events are
still programmed on the termination?

If so, the strict/exact and strict/state won't be of any help I'm
afraid:
the Modify will always overwrite events that were present on the termination
before... and if the MGC just sent this Modify to know whether the hook
status of a termination is still as it expected, without taking anything
else, this is a problem. Do I interpret this correctly?

Pascal


----- Original Message -----
From: "Kevin Boyle" <kboyle at nortelnetworks.com>
To: "Pascal Lambers" <pascal.lambers at pandora.be>; <megaco at ietf.org>
Sent: Tuesday, August 17, 2004 10:45 PM
Subject: RE: [Megaco] Modify to do Audit hook state


> The MGC must account for these kinds of issues when it is building
messages.
>
> I would note that there are two other possible values for the strict
parm
> that may be of more use: 'state' and 'exact'.  If having the MG
generate
> errors is a problem, then you could use one of the other two options 
> instead.
>
> Kevin
>
>
>
>   _____
>
> From: Pascal Lambers [mailto:pascal.lambers at pandora.be]
> Sent: Tuesday, August 17, 2004 4:32 PM
> To: megaco at ietf.org
> Subject: [Megaco] Modify to do Audit hook state
>
>
> Hello,
>
> Megaco does not provide an Audit message to audit a termination's hook 
> status directly. That's clear.
>
> Instead, to know whether a termination is Off Hook, the MGC must send
a
> Modify with an Off Hook (strict/failWrong) Events descriptor.
Consequently,
> if the termination is already off hook, the MG will reply with an
error.
Is
> it sure that in such a case the events and signals implemented on this 
> termination before the Modify arrived, are still present? In other
words,
is
> it sure that messages from the MGC to the MG replied to with an error
do
not
> have any negative effect?
>
> Kind regards,
> Pascal
>
>
>
>
> _______________________________________________
> Megaco mailing list
> Megaco at ietf.org
> https://www1.ietf.org/mailman/listinfo/megaco
>


_______________________________________________
Megaco mailing list
Megaco at ietf.org
https://www1.ietf.org/mailman/listinfo/megaco

_______________________________________________
Megaco mailing list
Megaco at ietf.org
https://www1.ietf.org/mailman/listinfo/megaco
_______________________________________________
Megaco mailing list
Megaco at ietf.org
https://www.ietf.org/mailman/listinfo/megaco