Re: SQL Question

["Joe Sam Shirah" <jshirah@xxxxxxxxxxxxx>]

    Hi Larry,

    You shouldn't have any problem if you use a standard forward-only
ResultSet.  The values are not accessible once you move to the next row and
should be available for garbage collection with a decent driver ( I'm
generalizing here. )

    There is a default fetch size, which determines the number of rows
downloaded per trip to the database; you can change it to suit your needs.
Use Statement.getFetchSize() and Statement.setFetchSize() for control.

    There are also other ways to connect, grab a set, disconnect, process,
repeat, but unless connections are scarce, you probably don't need to go
that route.


                                                         Joe Sam

Joe Sam Shirah -        http://www.conceptgo.com
conceptGO       -        Consulting/Development/Outsourcing
Java Filter Forum:       http://www.ibm.com/developerworks/java/
Just the JDBC FAQs: http://www.jguru.com/faq/JDBC
Going International?    http://www.jguru.com/faq/I18N
Que Java400?            http://www.jguru.com/faq/Java400


----- Original Message ----- 
From: "Larry" <larryhytail@xxxxxxxxx>
To: <java400-l@xxxxxxxxxxxx>
Sent: Tuesday, September 13, 2005 7:27 PM
Subject: SQL Question


>
> I want to do an SQL in a java program (JDBC).  It's something I'd run once
a month, but I know it will generate a lot of records in the record set.  If
I just did the query in one shot, I'd probably freeze my computer due to
insufficient memory.
>
> Is there a way to just bring down 1000 or so records at a time with each
repeated call of the SQL getting the next 1000 and etc until I hit the end
of what would have been one big result set?
>
> The method that lets you do this, is it efficient for the backend dataase
(iSeries), or does it actually run the whole query each time, but just
return the 1000 records you need for that call?
>
> Any advice?
>
> Thanks,
>
> Larry
>