Re: Dynamic SQL for multiple data libraries

On Wed, May 11, 2016 at 6:07 PM, CRPence <crpbottle@xxxxxxxxx> wrote:

Note: I see in followup replies, that the issue dlclark had resolved by
switching to dynamic, was a functional issue; i.e. the switch was made, to
circumvent a defect, not to improve performance.

​Well as I read it, the switch was made to circumvent a perceived defect
that simply left open data paths open and didn't appear to result in
incorrect output.

But thanks for the insight Chuck.

I hadn't thought about the cost to update the plan info in the *PGM/*SRVPGM
objects.

One question, as I understand it the system plan cache gets updated for
static statements. So after 600 runs, it would seem that the system plan
cache would be in the same state regardless of static or dynamic.

When the static statement is run again, would the system check in the plan
cache and make use of that plan (possibly updating the *PGM/*SRVPGM object)
or would it automatically do a full re-plan?

Charles