Re: CQE, SQE and QAQQINI

wrote:

I know that Infor recommend changing the IBM i 6.1 setting for
IGNORE_DERIVED_INDEXES back to *NO for BPCS / ERP LX installations. For
these, the majority (75% +) of the indexes are logical files with
Select/Omit criteria. Ignoring all the those indexes results in a slower
sytem overall, even though the SQE engine is faster.


________________________________

From: midrange-l-bounces@xxxxxxxxxxxx on behalf of Jonathan Mason
Sent: Fri 28/08/2009 13:53
To: midrange-l@xxxxxxxxxxxx
Subject: CQE, SQE and QAQQINI



We have an SQL statement that uses the CQE and runs extremely slowly. In
test I have created an Encoded Vector Index and also changed a local version
of the QAQQINI file so that the IGNORE_DERIVED_INDEXES is set to *YES.

With these settings the SQL runs extremely fast, however the powers that be
are nervous about changing a global setting that would potentially affect
all SQL requests.

I've spent the morning talking to my friend, Google, and searching through
the Midrange.com archives, and everything I have seen indicates that using
the SQE is better than using the CQE.

With this in mind, are there any pitfalls in changing the
IGNORE_DERIVED_INDEXES setting to *YES? Should we expect all SQL to run at
least as fast as does currently or is there a risk that some SQL processes
could run slower?

Thanks in advance

Jonathan



--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.



--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.