RE: Implement UNLOGGED clause for COPY FROM

From: "osumi(dot)takamichi(at)fujitsu(dot)com" <osumi(dot)takamichi(at)fujitsu(dot)com>
To: 'Amit Kapila' <amit(dot)kapila16(at)gmail(dot)com>
Cc: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, "David G(dot) Johnston" <david(dot)g(dot)johnston(at)gmail(dot)com>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>
Subject: RE: Implement UNLOGGED clause for COPY FROM
Date: 2020-07-22 05:41:15
Message-ID: OSBPR01MB4888CBD08DDF73721C18D2C0ED790@OSBPR01MB4888.jpnprd01.prod.outlook.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi. Amit-san

> If you are going to suggest users not to replicate such tables then why can't you
> suggest them to create such tables as UNLOGGED in the first place? Another
> idea could be that you create an 'unlogged'
> table, copy the data to it. Then perform Alter Table .. SET Logged and attach it to
> the main table. I think for this you need the main table to be partitioned but I
> think if that is possible then it might be better than all the hacking you are
> proposing to do in the server for this special operation.
Thank you for your comment.

At the beginning, I should have mentioned this function was
for data warehouse, where you need to load large amounts of data
in the shortest amount of time.
Sorry for my bad explanation.

Based on the fact that data warehouse cannot be separated from
usage of applications like B.I. tool in general,
we cannot define unlogged table at the beginning easily.
Basically, such tools don't support to define unlogged table as far as I know.

And if you want to do so, you need *modification or fix of existing application*
which is implemented by a third party and commercially available for data analytics.
In other words, to make CREATE UNLOGGED TABLE available in that application,
you must revise the product's source code of the application directly,
which is an act to invalidate the warranty from the software company of B.I. tool.
In my opinion, it would be like unrealistic for everyone to do so.

Best,
Takamichi Osumi

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Amit Langote 2020-07-22 06:16:02 Re: problem with RETURNING and update row movement
Previous Message Michael Paquier 2020-07-22 05:35:10 Re: OpenSSL randomness seeding