Re: Way to get STRSQL SELECT to run in batch?

Dan

You can do a couple things to investigate this.

1. STRDBG in both jobs - you'll need to use STRSRVJOB fisrt on the batch one - or use service entry point maybe. Then in the job log will be messages about what the optimizer chose to do - it's probably different for each one - the interactive version does have a different assumption about resources and all.
2. STRDBMON before running each one, the ENDDBMON afterwards and put the results in a PF - it's an ugly beast but can be imported into Visual Explain in Navigator, then you get lots of cool information. You CAN also make monitors in Navigator that can work over different combinations of jobs.

HTH
Vern

-------------- Original message --------------
From: Dan <dan27649@xxxxxxxxx>

On Tue, Jul 29, 2008 at 12:49 PM, Elvis Budimlic <
ebudimlic@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:

Build 2 indexes (i.e. CREATE INDEX) if you don't already have keyed lfs
over
these fields:

1) CVCUSN, CVRENN, CVSEQN
2) cdCUSN, cdRENN, cdSEQN, cdstat

The indexes above ought to help the performance of your query.

Elvis, the table for 1) has as its primary key those three fields, followed
by three other fields. The table for 2) has as its primary key the first
three field, but not cdstat.

What is most surprising, as I mentioned in my response to Eric, is that the
interactive query took hours to build an index on the 1) table, but when I
submitted it to batch per the CREATE TABLE ZZZZ AS... it took less than 5
seconds for it to run! What gives?

Really mind boggling.

- Dan
--
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.