Re: Hint is not used in SQL
Date: Wed, 15 Nov 2023 10:51:58 -0500
Message-ID: <CAG67e6SGiHE4VvjLa17q_erRWEmjvioNW+ikNJv4gMiHwHpJWg_at_mail.gmail.com>
Hello Jonathan,
My fault I believe I messed it up when taking the tests. I dropped and
reapplied the patch, and I generated the expanded hint report once more.
Because I know we are only changing one row from the front end, the view
must run first, followed by the table.
I'm not sure if VIEW PUSH PREDICATE is causing the leading hint to fail,
but it also didn't work when I used NO PUSH PREDICATE even though it was
used but LEADING was not.
Best Regards,
> That hint report and patch generation statement aren't consistent with the
AMIT
On Wed, Nov 15, 2023 at 10:40 AM Jonathan Lewis <jlewisoracle_at_gmail.com>
wrote:
> previous output.
>
>
>
>
>
>
> *3 - SEL$AC90CD92 / PO_LINE_LOCATIONS_ALL_at_SEL$2 U -
> USE_NL_WITH_INDEX(_at_SEL$AC90CD92 PO_LINE_LOCATIONS_ALL_at_SEL$2) 5 -
> SET$632CC7ED N - LEADING(_at_SET$632CC7ED VW_NSO_1_at_SEL$AC90CD92
> PO_LINE_LOCATIONS_ALL_at_SEL$2) *
>
> You have *_at_SET$632CC7ED* in the leading hint and that should be
> *_at_SEL$AC90CD92. *
> For an initial test of the patch I'd also use just use_nl() rather than
> use_nl_with_index().
>
> Regards
> Jonathan Lewis
>
>
>
-- http://www.freelists.org/webpage/oracle-lReceived on Wed Nov 15 2023 - 16:51:58 CET
- text/plain attachment: hint_report.txt