[OpenSIPS-Users] newtran() limitations

[Bogdan-Andrei Iancu]

Hi Brett,

if you create sooner the transaction, you have bigger chances to catch a 
retransmission asap - detecting retransmission asap helps you because 
you avoid any processing (in script) for a retransmission.

Regards,
Bogdan

Brett Nemeroff wrote:
> Bogdan,
> This kind of explanation is really helpful. If you don't mind, I'd 
> like to ask a little more detail.
>
> Can you please explain when it would be necessary to force creation of 
> a transaction where t_relay wouldn't suffice?
>
> Thanks,
> Brett
>
>
> On Wed, Oct 7, 2009 at 9:50 AM, Bogdan-Andrei Iancu 
> <bogdan at voice-system.ro <mailto:bogdan at voice-system.ro>> wrote:
>
>     Hi Stan,
>
>     Just to explain you want actually happens when you do t_newtran() :
>
>     This function forces the creation of the transaction (which
>     otherwise is
>     automatically done by functions like t_relay() and t_reply() ).
>     Part of
>     creating the transaction is cloning of the current request into shared
>     memory (cloned from pkg into shm). We need this copy in
>     transaction (is
>     shared mem) as the processing of this request may continue in a
>     failure
>     route (somewhere later, in other process).
>
>     So, the warning is about this cloning/copy action - the request, as it
>     is at that moment, will be copied into shm for later usage -
>     whenever/whenever you do use later the transaction, you will see the
>     message (and its changes) from the cloning time. Changes that are
>     copies
>     at cloning time are the lumps (changes on the textual level). Other
>     things like flags, RURI are updated in transaction at t_relay() time.
>
>     Hope this story will answer your questions :)
>
>     Regards,
>     Bogdan
>
>     Stanisław Pitucha wrote:
>     > Hi,
>     >
>     > I'd like to know what exactly does this note mean: (from the
>     documentation)
>     >
>     > "NOTE that the changes on the request that are made after this
>     > function call will not be saved into transaction!!!"
>     >
>     > What is the limitation in practice / what is saved into transaction.
>     > In details - if I have an INVITE packet and call t_newtran() on it:
>     > - will modification of the message body take effect on t_relay()?
>     > (done both before and after newtran)
>     > - will modification of $ru and $du have any effect after newtran()
>     > - what other scenarios should I think about that might be affected?
>     >
>     >
>
>
>     _______________________________________________
>     Users mailing list
>     Users at lists.opensips.org <mailto:Users at lists.opensips.org>
>     http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Users mailing list
> Users at lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>