Re: Index Selection

Thanks Charles...I'll examine the data. I removed a DDM file that was used
on system B, and that increased the performance. I'll check out the QAQQINI
files too.


On Fri, Aug 9, 2013 at 3:36 PM, Charles Wilt <charles.wilt@xxxxxxxxx>
wrote:

What's the data look like between A & B?

Same total number of rows?

Same number of rows returned?

Recall that the system will basically always use arrival sequence if it
thinks more than xx% of the total will be returned... xx% is a

relatively

low number, though I can't find it right now. I'm pretty sure it's less
than 30..actually 10% sticks in my head..

You might also check that the QAQQINI file in effect for the query is the
same on both systems.

Charles


On Fri, Aug 9, 2013 at 3:20 PM, Michael Ryan <michaelrtr@xxxxxxxxx>

wrote:

Yes, B's performance is much poorer. Like a 2 second display build on A

vs.

a 10+ second display build on B


On Fri, Aug 9, 2013 at 3:08 PM, Charles Wilt <charles.wilt@xxxxxxxxx>
wrote:

Run Visual Explain from iNav's Run SQL scripts...

But I'd be willing to bet the answer is "the cost to use this index

was

to

high" or whatever the exact wording is.

Unless the systems are configured exactly the same, the query is

running

over the exact same data _AND_ the systems are running the exact same
workload at the time the query is done, there's no reason to assume

the

SQE

would chose to use the exact same method.

Is the performance of system b significantly poorer?

Charles



On Fri, Aug 9, 2013 at 2:56 PM, Michael Ryan <michaelrtr@xxxxxxxxx>

wrote:

Ok...two systems, both on 7.1. Execute STRDBG UPDPROD(*YES). Run a

program

on system A. Job log shows an index was used for bit map

processing.

Run

the same program on system B. Same index in place (though different
libraries). System B job log says arrival sequence was used. How

can

I

figure out why the index is used on one system and not the other?

Thanks!

--
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.

--
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.

--
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.