20 Jan
2003
20 Jan
'03
4:04 p.m.
Maxim,
I think that's a good plan.
Well -- I'm hearing to embarassement of folks to whom I promised we will
never ever even think of b2bua. Let me explain.
I think, that architecturally, the concept of B2BUA is a departure from
Internet's end-2-end model. As some other lessons like NAT show, that
may have very inconvenient consequences in course of the time, even
such we do not predict now (like single point of failure and scalability
issues).
One can achieve complex services without any such B2BUA with a more
distributed design -- end-devices just need to be smart enough and
participate in service composition. For example, session-related policy
like maximum call duration and keeping firewalls open can be achieved
using session-timer. Multi-phase conversation with multiple components
may be implemented using call transfer (anouncement component hands over
to PIN-collecting component which hands over to gateway, etc.) I hope
I will be able to give some more details in a memo I'm writing,
Rohan Mahy's cc-framework Internet draft spells some issues very well.
The problem is that many available components (like PSTN gateways) are
quite dumb and do not implement all needed pieces yet. For example,
a gateway is not able to transfer to an anouncemnt server ("call has
een cut-off, give me your credit card number") when a session timer hits.
To utilize such components for service composition, B2BUA is the only way
to move forward.
Jan is right, that there are some issues with ser -- we want to keep it
reasonably compact and performance has not been optimized for maintenance
of call state. However, as long as the B2BUA is kept in a separate module
tied to TM via callbacks, it will be still useful, whereas users wishing
to use compact ser are no way forced to use it.
Just ask us if you need any guidance on TM or whatever.
-Jiri
At 10:37 PM 1/16/2003, Maxim Sobolev wrote:
Hi,
I am thinking about writing a new module for SER, which will track SIP
dialogs and will serve as an abstraction layer for other modules, much
like the tm module now. We need such module for 2 reasons:
1. Call accounting. Our billing engine is based on the assumption that
a node provides accounting information for completed calls, not for
individual transactions. It is easier for us to extend proxy with
similar features than to modify billing engine to do transaction
matching.
2. Debit card application. Currently, there is no way to use SER for
debit card applications, where it is necessary to set the maximum
duration of the call and terminate it forcefully if that duration is
exceeded.
The raw idea is as follows:
- the module will register callbacks with tm.register_tmcb(), probably
TMCB_REQUEST_IN and TMCB_REPLY_IN ones and will match INVITEs to BYEs
keeping information about the state of ongoing sessions in the shared
memory.
- the module will provide interested modules with ability to register
several callbacks, i.e. on dialog creation/teardown and yet another
callback on dialog timeouts (more about that below).
- the module will provide utility functions for forceful termination
of any ongoing dialog.
- when invoking dialog creation callback function, the module will
give the function opportunity to install a timer on that dialog, so
that if the dialog is still active after timer expires, then some
action is performed. For example, in the debit card applications, in
such case the accounting module can decide to forcefully terminate a
dialog.
What do you think?
-Maxim
_______________________________________________
Serusers mailing list
serusers(a)lists.iptel.org
http://lists.iptel.org/mailman/listinfo/serusers
--
Jiri Kuthan http://iptel.org/~jiri/