Alexander Malysh wrote:

> I would like to propose a simple and "well-known" procedure to generate
> "unique" id for network applications:
>         1) id = official ip + pid (process id) + timestamp (sec. resolution enough) +
> global counter
>         2) id = pid + timestamp + global counter
> 
> 1) is unique for all running bearerboxes at least in your private ip address
> range.
> Descr.: Official ip is a unique idendifier against other hosts, pid +
> timestamp protect from a case if bearerbox crached too oft...
> 
> 2) is unique for all running bearerboxes on the same machine.
> 
> I would preffer 1) so we never have any problems with every possible
> configuration...

sounds like you pulled some RFC out of your hands?! ;) any references
on this "well-known" procedure? 

or is it well-know by you?! :p

BTW, alternative 1) sounds reasonable for me.

How about propagation of the unique ID creation mechanism for back-end
systems (ie. application server) where MT messages are created and
transported via Kannel?

Should the back-end system log such IDs to database for tracing, or
should it pass the message and expect Kannel to deliver the unique ID
as a response?

We may have this substitute the current "Sent." response in smsbox, by
letting the response thread wait until bearerbox sends the ACK for the
Msg and literally pass the sms.id to the HTTP client instead simply
saying "Sent."?!

Stipe

[EMAIL PROTECTED]
-------------------------------------------------------------------
Wapme Systems AG

Vogelsanger Weg 80
40470 D�sseldorf

Tel: +49-211-74845-0
Fax: +49-211-74845-299

E-Mail: [EMAIL PROTECTED]
Internet: http://www.wapme-systems.de
-------------------------------------------------------------------
wapme.net - wherever you are

Reply via email to