RE: OPNQRYF after 5.1

["Andy Nolen-Parkhouse" <aparkhouse@xxxxxxxxx>]

Kirk,

Here are some of my questions/thoughts:

Is this all on the same AS/400?

What is a location?  If these locations are all accessing the same
AS/400 via some kind of green-screen, then where they are located should
make no difference.  Look for differences in user profiles, library
lists, authorities, etc.

If you display a slow and a fast job, what are the differences?

Do 'locations' deal with different library lists which would have the
same query run over different DB files?  A trigger would relate to the
database, not the program and could account for the differences.

Triggers do not magically appear on user database files after an
upgrade.  Is this a shop in which someone may have experimented with new
features after the upgrade?  (This assumes they are running against
different files).

Regards,
Andy Nolen-Parkhouse

> On Behalf Of Kirk Goins
> Subject: OPNQRYF after 5.1
>
> I've got a client that has an OPNQRYF that at v4r5 was approx 1sec for
all
> users.
> After upgrading to v5r1 selected offices are getting 3-4sec response
with
> the
> rest getting approx 1sec response. The programmer says same program in
all
> cases...
>
> I did a strdbmon *detail while the programmer ran it as a fast
location
> and then
> ended and restarted dbmon and captured the data with the programmer
> running like
> a slow office.
>
> What I found:
> #1 excluding the start/stop dbmon cmd the fast office captured 3 lines
> #2 excluding the start/stop dbmon cmd the SLOW office captured 18
lines
> #3the slow office has 15s line dealing with TRIGGERS, then the OPNQRYF
> same as
> the fast office.
>
> Monday we are going run it in debug and see what we can see but I
would
> like
> some more ammo as well. I know what triggers are and basically what
they
> are
> used for.
> How would you go about tracking down this problem?