| From: | Christophe Pettus <xof(at)thebuild(dot)com> |
|---|---|
| To: | Pgsql-performance <pgsql-performance(at)postgresql(dot)org> |
| Cc: | Nagaraj Raj <nagaraj(dot)sf(at)yahoo(dot)com>, David Rowley <dgrowleyml(at)gmail(dot)com> |
| Subject: | Re: issue partition scan |
| Date: | 2021-05-26 00:27:27 |
| Message-ID: | 566EF81B-F9B2-4CD2-900E-5163732770FF@thebuild.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
> On May 25, 2021, at 17:16, David Rowley <dgrowleyml(at)gmail(dot)com> wrote:
>
> It's because of the OR condition. If it was an AND condition then the
> planner wouldn't have to consider the fact that records in other
> partitions might be required for the join.
The OP might consider rewriting the query as a UNION, with each part of the top-lkevel OR being a branch of the UNION, but excluding the partitioned table from the JOINs for the branch of the UNION that doesn't appear to actually require them.
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Eugen Konkov | 2021-05-27 10:37:33 | Count (select 1) subquery as constant |
| Previous Message | David Rowley | 2021-05-26 00:16:58 | Re: issue partition scan |