Closing ADODB sub reference outs in VIs? (LV 8.5.1)

Question

I had a question about closing references to items I have not explicitly opened.
e.g.
I create an ADODB connection object.
From that, I create an ADODB command object.
One of the Command properties is the Parameters reference. I do a parameters.Refresh and using count, I loop through a For loop getting Item references. For each Item reference, I get a parameter object.
One of the properties in the parameter object is a Properties reference.  I do a refresh on that and then go into another For loop. For each Item reference, I get a property node.
OK, now that I've gone through getting/setting information, I need to start working my way out. The last property node had an item reference out from the properties node. Should I close that? The properties node itself had as a reference out, should that be closed? The Parameter node has a reference out, should that be closed? Finally, the Parameters node has a reference out, should I close that?
The command and connection nodes I explicitly opened, so I know I need to close them. But these other references, should I be closing them, or do I only need to close the references I explicitly created?
Thanks, Bob

Answer 1

"intp56" <[email protected]> wrote in message
news:[email protected]..
> I download LabSQL, it looks pretty close to what I want.
> &nbsp;
I'd suggest to use it, and to send them your additional code in return if
you're alowed to do that... Or suggest what you are missing to the project,
so they can add it in later versions.
> I'm looking for a VI that queries a stored procedure to determine what the
parameters are, then creates a LabVIEW cluster with the correct datatypes.
The idea is to call that first to create the Parameters object, then use it
later, particularly for DML operations. (Say insert rows from a LV table.)
I've tried using ADO recordsets to insert large datasets, and found it's
performance to be lacking or unacceptable. I can insert several 1000 records
per second with successive calls to a stored procedure where I specify ahead
of time no result set.
Not sure how ADO does this. I vaguely recall something about it. I think the
values are send item by item, and I think there was a faster way to send the
data by using a variant that directly accepts an array... It's been a while
since I've messed around with that, and I might not have succeeded to get it
working. Googling for "ADO large data sets" or something might give some
clues.
> If I really have lots of rows to insert, then I've found it better to
prepare a file, move it to a place where the database can see it, and BULK
INSERT it asynchronously and report the results back later.
Can't imagine that this is the best way to do it, but it's a good
workaround.
> BTW, I liked the tip that any reference WIRE needs to be closed, that
gives me a nice rule to know what I need to closed.
But even if there is no wire comming out of a reference property, you still
need to close it. So: every "green" read property should be closed.
Regards,
Wiebe.