Date: Thu, 15 Jul 2021 22:53:12 -0400
From: Alvaro Herrera <alvherre@alvh.no-ip.org>
To: Ron <ronljohnsonjr@gmail.com>
Cc: pgsql-admin@lists.postgresql.org
Subject: Re: Partitioning on the date part of a timestamp & PK issues
Message-ID: <202107160253.va6esfntkuxr@alvherre.pgsql>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <60a7d2e0-787e-68fe-28d6-40ec32d11c0f@gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/202107160253.va6esfntkuxr%40alvherre.pgsql>
Precedence: bulk

On 2021-Jul-15, Ron wrote:

> CREATE TABLE t (
>     guid uuid not null,
>     seq smallint not null,
>     tid smallint not null,
>     ts timestamp without time zone not null,
>     x real,
>     y real,
>     z real,
>     primary key (guid, tid, seq, ts)
> )   partition by range (ts);

This definition will let you have multiple rows for the same
(guid,tid,seq) combination, as long as they have different ts values.
That is at odds with the original formulation, in which there could be a
single row with any given combination of those three columns.

-- 
Álvaro Herrera         PostgreSQL Developer  —  https://www.EnterpriseDB.com/
"Having your biases confirmed independently is how scientific progress is
made, and hence made our great society what it is today" (Mary Gardiner)