Re: Tuning odbc query for performance

Jay

The help text in ACS for the visual explain  utility is really quite good, I suggest going there.

You will want to get familiar with the different information you get in the right-hand panel (IIRC) = things like why it chose a certain index to use.

There are also index recommendations that you may or may not want to use.

Finally, if this runs in production, you can find it as one of the SQL statements for a job - this will give you a better sense of things, because it was done in the production environment - memory, job contention, all manner of things can be very different between running something in ACS and running in production.

And if you can, come to COMMON, there will be, as I recall - there's one session on revving up SQL performance with indexes. - you might be able to find that info elsewhere on the 'net, too.

HTH
Vern

On 5/2/2019 4:49 PM, Jay Vaughn wrote:

So... we have a fairly large (not complex) sql statement we are calling via odbc.

Select = 200 columns straight from tables. No functions

From = primary table and 12 left joins all on loan number key on each table.

Where = loannumber = ?

Order by = various columns from the tables in desc

So the query is pretty vanilla with not a lot of opportunity for tweaking for performance. What I’m interested in is anything that can be done within its “run time environment”?
I did a run/explain on the statement in ACS and it provided a ton of info I’m guessing that correlate mostly to the query optimizer. I would think if I had some good knowledge on what any of them are and how to adjust per query execution then I’d be headed in the right direction.

Can anyone offer any suggestions?

Tia

Jay