From: | Сергей _ <argonaut(dot)od(at)gmail(dot)com> |
---|---|
To: | pgsql-general(at)lists(dot)postgresql(dot)org |
Subject: | Reference-Partitioned Tables |
Date: | 2020-11-06 15:56:13 |
Message-ID: | CANJ_Omo0zW04f6ODOBU2_3XG9UyOr2zz0W48kJFy3kCR4YOymA@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Hi,
There is a proposal to include in future developments sectioning by
reference, as in Oracle
https://docs.oracle.com/database/121/VLDBG/GUID-00923EB3-05F6-41F7-8437-E42FC9BD9571.htm#VLDBG1093
There is a widespread model when data related to properties of a superclass
is stored in a base table, and data related to subclasses in a detail table. Or
when the main entity has volatile additional parameters. A very simplified
example in the attached image.
[image: contracts.jpg]
The most successful choice of partitioning table *Contracts* is sectioning
by list with key column *ContractTypeID*. It would be nice if the table
*ContractDetails* is partitioned automatically like the parent table
*Contracts*. Then we don't have to add a crutch column *ContractTypeID* to
the table* Contract**Details* and partition manually in sync with
*Contracts*. This field also consumes disk space, since the table of details
is usually large and there can be more than one partitioning key.
Thanks, best wishes.
From | Date | Subject | |
---|---|---|---|
Next Message | Gabi Draghici | 2020-11-06 16:12:40 | Re: pgagent |
Previous Message | Adrian Klaver | 2020-11-06 15:38:28 | Re: pgagent |