Re: [reSIProcate] Removing Transport
Hi,
Ok ... its not about deleting transport, but shutting-down the transport
as Byron mentioned precisely. Yeah there would be repeated re-configuration,
but that is the intent for shutting down the transport and adding a new as per the
new configuration. It would be a pain to remove the SIP stack object and DUM
object every time I want to modify (closing down older one, and adding new transport)
transport. In that sense it is surely a desirable stuff!
I understand doing it with handles will really require a LOT of work, but can we
consider failing the DUM operation if the asked transport is not found at transport.
I feel that there should not be a state-bloat in this case at DUM.
I am not pretty sure of this theory, but would really appreciate if you can give some
thought and validate it.
Thanks for your time :)
Nilay
On 12/9/06,
Matthias Moetje <moetje@xxxxxxxxxxxx> wrote:
Byron,
I would also be very intered in the ability to
dynamically remove transports. My opinion though,
is that invalidating transports without deletion
would be just fine. Changing transports is a task
that is usually done rarely. Some dead transports
flying around wouldn't do any harm.
Sure, there may be rare cases when someone might
want to do this frequently, but would these rare
cases be worth to undertake so much effort? I think
the "shut down" solution would be fine for most.
What do others think?
Best regards
Matthias Moetje
______________________________________________
TERASENS GmbH Phone: +49.89.143370-0
Augustenstraße 24 Fax: +49.89.143370-22
80333 Munich e-mail: info@xxxxxxxxxxxx
GERMANY Web: www.terasens.com
______________________________________________
> -----Original Message-----
> From: resiprocate-devel-bounces@xxxxxxxxxxxxxxxxxxxx
> [mailto:resiprocate-devel-bounces@xxxxxxxxxxxxxxxxxxxx] On
> Behalf Of Byron Campen
> Sent: Friday, December 08, 2006 3:37 PM
> To: Nilay Tripathi
> Cc: resiprocate-devel@xxxxxxxxxxxxxxxxxxxx
> Subject: Re: [reSIProcate] Removing Transport
>
> Currently there is no way to remove transports after
> they have been added.
>
> Allowing for the removal of a transport could end up
> being very tricky. There are lots and lots of Transport*
> (mostly in Tuple) floating around, and deleting a Transport
> will almost certainly invalidate lots of them. If we were to
> "shut down" transports (without actually deleting them), this
> would be safer, but we could end up with state-bloat after
> repeated re-configuration. Removing all of these pointers and
> replacing them with some sort of handle could work, but a lot
> of code would have to be touched.
>
> Mind you, this is still something I eventually intend
> to do, but I have a lot on my plate to finish first.
>
> Best regards,
> Byron Campen
>
> > Hi All,
> >
> > I was looking for a way to remove an added transport at the
> > application.
> > Actually, I want to be able to either remove or update a transport
> > within the application, so that I do not have to close it down.
> >
> > It seems some api like removeTransport( ) was there in DUM
> till 0.9.0.
> > (I am working at 1.0 now) which is not there in new
> versions. And as
> > per my understanding goes, the transport is finally removed
> from the
> > map at TUSelector when SipStack destructor is called during
> shutdown.
> >
> > Please throw some light, if there is any other way that I can
> > remove/update transport or will I have to close down SIP
> Stack and DUM
> > each time to do it.
> >
> > Thanks,
> > Nilay
> > _______________________________________________
> > resiprocate-devel mailing list
> > resiprocate-devel@xxxxxxxxxxxxxxxxxxxx
> > https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>
> _______________________________________________
> resiprocate-devel mailing list
> resiprocate-devel@xxxxxxxxxxxxxxxxxxxx
>
https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>