| From: | Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org> |
|---|---|
| To: | Pavan Deolasee <pavan(dot)deolasee(at)gmail(dot)com> |
| Cc: | Andres Freund <andres(at)anarazel(dot)de>, Robert Haas <robertmhaas(at)gmail(dot)com>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, amul sul <sulamul(at)gmail(dot)com>, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com>, Stephen Frost <sfrost(at)snowman(dot)net>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: [HACKERS] Restrict concurrent update/delete with UPDATE of partition key |
| Date: | 2018-04-05 19:34:54 |
| Message-ID: | 20180405193454.el2bffwijtsa7rbg@alvherre.pgsql |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Pavan Deolasee wrote:
> On Thu, Apr 5, 2018 at 7:14 AM, Andres Freund <andres(at)anarazel(dot)de> wrote:
> + /*
> + * As long as we don't support an UPDATE of INSERT ON CONFLICT for
> + * a partitioned table we shouldn't reach to a case where tuple to
> + * be lock is moved to another partition due to concurrent update
> + * of the partition key.
> + */
> + Assert(!ItemPointerIndicatesMovedPartitions(&hufd.ctid));
> +
>
> This is no longer true; at least not entirely. We still don't support ON
> CONFLICT DO UPDATE to move a row to a different partition, but otherwise it
> works now. See 555ee77a9668e3f1b03307055b5027e13bf1a715.
Right. So I think the assert() should remain, but the comment should
say "As long as we don't update moving a tuple to a different partition
during INSERT ON CONFLICT DO UPDATE on a partitioned table, ..."
FWIW I think the code flow is easier to read with the renamed macros.
--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Andres Freund | 2018-04-05 19:43:15 | Re: [HACKERS] Restrict concurrent update/delete with UPDATE of partition key |
| Previous Message | Bruce Momjian | 2018-04-05 19:33:14 | Re: PostgreSQL's handling of fsync() errors is unsafe and risks data loss at least on XFS |