Horton

Re: pg_column_toast_chunk_id: a function to get a chunk ID of a TOASTed value

Yugo Nagata <nagata@sraoss.co.jp>

View thread
From: Yugo NAGATA <nagata@sraoss.co.jp>
To: Nathan Bossart <nathandbossart@gmail.com>
Cc: jian he <jian.universality@gmail.com>, Sergei Kornilov <sk@zsrv.org>, Nikita Malakhov <hukutoc@gmail.com>, pgsql-hackers@postgresql.org
Date: 2024-03-12T08:51:19Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Add pg_column_toast_chunk_id().

Attachments

On Sat, 9 Mar 2024 08:50:28 -0600
Nathan Bossart <nathandbossart@gmail.com> wrote:

> On Sat, Mar 09, 2024 at 11:57:18AM +0900, Yugo NAGATA wrote:
> > On Fri, 8 Mar 2024 16:17:58 -0600
> > Nathan Bossart <nathandbossart@gmail.com> wrote:
> >> Is this guaranteed to be TOASTed for all possible page sizes?
> > 
> > Should we use block_size?
> > 
> >  SHOW block_size \gset
> >  INSERT INTO test_chunk_id(v1,v2)
> >   VALUES (repeat('x', 1), repeat('x', (:block_size / 4)));
> > 
> > I think this will work in various page sizes. 
> 
> WFM
> 
> > +SHOW block_size; \gset
> > + block_size 
> > +------------
> > + 8192
> > +(1 row)
> 
> I think we need to remove the ';' so that the output of the query is not
> saved in the ".out" file.  With that change, this test passes when Postgres
> is built with --with-blocksize=32.  However, many other unrelated tests
> begin failing, so I guess this fix isn't tremendously important.

I rewrote the patch to use current_setting('block_size') instead of SHOW
and \gset as other tests do. Although some tests are failing with block_size=32,
I wonder it is a bit better to use "block_size" instead of the constant
to make the test more general to some extent.

Regards,
Yugo Nagata

> 
> -- 
> Nathan Bossart
> Amazon Web Services: https://aws.amazon.com


-- 
Yugo NAGATA <nagata@sraoss.co.jp>