when you create the trace, in the Filter Tab select DatabaseID Like "Yourdatabase ID"

AND check Exclude system IDs

just keep in mind that if the proc already perform insert ... exec in it you won't be able to do it from the outside (nesting)

what do you mean by "state"

I think you already answer your own question:

create function dbo.udf_concat_products (@Center int )

returs varchar(400)

as

begin

 declare @str varchar(400), @delimiter char(1)

 SET @delimiter = ','

 SELECT @str =COALESCE(@str + @delimiter, '') + Product

 FROM...

May I suggest that you compute the AVG at "INSERT Time" and then materialize those if the Number of AVG types is Fixed  

the thing with the queues is that traffic is still there and is another technology to the pile (SP, set up etc). Doable though!

In my last incursions in the SCADA...

Just hope you don't run that proc too often. Recompilation is CPU intensive when done frequently

I worked as a SCADA engineer a long time ago. Your solution is to transfer info the other way around.

1.The Pooling stations register with the server.

2.There is a trigger that...

Agreed!

OK... OK...

Create the job as you normally would

script the job out

in BOL look at the parameter:

@freq_subday_type  of  sp_add_jobschedule

the possible values are :

0x1

At the specified time

0x4 Minutes...

no prob

What happens is that it will wait for the next minute mark effectively running at 1min and 50sec

Isn't that, what I just said ?

There is an undocummented way of doing it  but why do you need a job to run that often?

I don't see how you use the inserted table (if you need it at all)

If you don't need that just run the sp directly

oh and I am also interested in...