RE: SQL issues in program

Could commitment control have anything to do with the halt? I don't use
*Commit so excuse me if that's a dumb question, but I thought that somewhere
in the program one had to COMMIT the backlog.

Jerry C. Adams
IBM i Programmer/Analyst
The value of history is not scientific but moral. By liberalizing the mind,
by deepening the sympathies, by fortifying the will, history enables to
control not society, but ourselves - a much more important thing; it
prepares us to live more humanely in the present and to meet rather than to
foretell the future. -Carl Becker
--
A&K Wholesale
Murfreesboro, TN
615-867-5070


-----Original Message-----
From: rpg400-l-bounces@xxxxxxxxxxxx [mailto:rpg400-l-bounces@xxxxxxxxxxxx]
On Behalf Of Charles Wilt
Sent: Thursday, September 22, 2011 7:44 AM
To: RPG programming on the IBM i / System i
Subject: Re: SQL issues in program

Pete,

This was a program using SQL?

I've seen FRCRATIO used incorrectly to solve the issue of blocked
writes when using RPG RLA, but AFAIK, SQL doesn't buffer like RPG
does.

Thus AFAIK, SQL I/O is always immediately passed to the DBMS and
always immediately available; regardless of rather or not the data has
been written to DASD.

Now-a-days FRCRATIO is usually the wrong answer....

Charles

On Wed, Sep 21, 2011 at 8:53 PM, Pete Hall <pete@xxxxxxxxxxxxxx> wrote:

On Tue, 20 Sep 2011 12:07:51 -0600, Scott and Kelly Curry wrote:

I have an SQLRPGLE program that first calls a procedure to insert
records into File A. I then have another procedure that is called
repeatedly for different files. This procedure uses dynamic SQL to
insert records into different files using certain criteria along with
File A. Then it uses dynamic SQL to delete the same records it inserted,
also using File A.

This is just a WAG, but I have an app that uses a work table in a similar
manner, to compose a document, process it, then delete the rows and start
on the next document. I had issues with the system caching the writes, so
that the data wasn't always there. I added an override like this:
       OVRDBF FILE(FOO) FRCRATIO(1) OVRSCOPE(something_appropriate)
That forces the system to write the data to disk immediately. The
downside is that it forces the system to write the data to disk
immediately. That's not efficient. But, it solved my problem, and it's
easy enough to try.

--
Pete Hall
pete@xxxxxxxxxxxxxx
--
This is the RPG programming on the IBM i / System i (RPG400-L) mailing

list

To post a message email: RPG400-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/rpg400-l
or email: RPG400-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/rpg400-l.