I agree with you that it would be better to have a background trigger
run by the database server. That was exactly what I asked. The answer
was that there is no current way to do a background trigger and to
accomplish what I want to do I need to write a daemon that calls Listen
and then on the Notify run my function.
If background triggers were a possibility that would make the whole
thing so much easier.
Rafal Pietrak wrote:
On Wed, 2006-05-24 at 13:36 +0200, Sim Zacks wrote:
My understanding of Listen/Notify is that it is a completely
disconnected process running on the database server.
Yes. But In my particular case (and I presume, the intention of
'bacground triggers' is that) a programmer (like myself) is not really
interested in the outcome of thet trigger process. So there is no
'technical' need for him/her to create the server side proces *provided*
hi/she can setup a job *within* the database server itself, and just go
away.
That's the idea of 'background triggers'.
Surely, there are work arounds. Like the LISTEN/NOTIFY server (not
datagase server, but system server) daemon that takes database server
notiffications. And even a system server daemon, that simply uses
synchronous database communication (like I did in my case). The problem
is, that I have this 'eatching desire', to have such technical issues
supported 'withing the framework of RDBM architecture'.
That's why I keep thinking, that the solution I used is actually
'bitting the fances', while gurus do it some other, *better* way. But if
not, a think that 'bakground triggers' could help here.
I may not have understood exactly what you are trying to do, but from
what I understood, this will solve your problem.
I think you did. I just feel that 'background triggers' is 'real life'
engineering issue, so it should get some backing from RDBMS.
just my 2c.
-R
---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster
