Re: Error related to hash memory

I've done a little check in the ORAUS.MSG file, and a couple of quick tests. The event that might help you pin down the problem is 32699; the level is a bitmap.

On a small test I got different outputs for levels 1,2,4,8,16 so you could try:

alter session set events '32699 trace name context forever, level 31';

you could try level 8191 in case there are other bit flags that do something with larger examples; my example wasn't spilling to disc.

Regards
Jonathan Lewis

On Thu, 2 Dec 2021 at 10:56, Jonathan Lewis <jlewisoracle_at_gmail.com> wrote:

>
> Just a little follow-up: I'd forgotten that I had published a note a few
> years ago about the "depth" error in execution plans with case/decode()
> that included embedded scalar subqueries; but I went looking for it and
> it's here ( https://jonathanlewis.wordpress.com/2014/10/19/plan-depth/ )
> if you want to check it out.
>
>
>

--
http://www.freelists.org/webpage/oracle-l