Re: Sql patch and driving_site hint

From: Jonathan Lewis <jlewisoracle_at_gmail.com>
Date: Sat, 8 May 2021 11:09:10 +0100
Message-ID: <CAGtsp8=1iaXyqfFvXdbkun3gdWJHgT783BRC7931h2uiTkYpJg_at_mail.gmail.com>



I should have said - if you want to get the right "alias_at_query_block" text then you could start by reporting the current execution plan with the ( format=>'alias' ) option to see the "Query Block / Object Alias" information, and that will give you the fully qualified name that you need to use.

Regards
Jonathan Lewis

On Sat, 8 May 2021 at 10:54, Jonathan Lewis <jlewisoracle_at_gmail.com> wrote:

>
> It works in my base test (on 11.2.0.4) - how are you speciying the hint?
>
> My example is very simple and if I edit the sql to add /*+
> driving_site(t2) */ then I get the required plan, but if I use an SQL patch
> to do this Oracle reports that it "used" the patch, but the plan doesn't
> show the remote as the driver. However, if I use the fully qualified alias
> /*+ driving_site(t2_at_sel$1) */ then the path is "used" and has the
> required effect.
>
> (The same was true for 19.3, so it's not an "old version" thing)
>
> Regards
> Jonathan Lewis
>
>
>
>

--
http://www.freelists.org/webpage/oracle-l
Received on Sat May 08 2021 - 12:09:10 CEST

Original text of this message