Re: Allow logical replication to copy tables in binary format

From: Amit Kapila <amit(dot)kapila16(at)gmail(dot)com>
To: Peter Smith <smithpb2250(at)gmail(dot)com>
Cc: Melih Mutlu <m(dot)melihmutlu(at)gmail(dot)com>, vignesh C <vignesh21(at)gmail(dot)com>, "Hayato Kuroda (Fujitsu)" <kuroda(dot)hayato(at)fujitsu(dot)com>, "shiy(dot)fnst(at)fujitsu(dot)com" <shiy(dot)fnst(at)fujitsu(dot)com>, Euler Taveira <euler(at)eulerto(dot)com>, Bharath Rupireddy <bharath(dot)rupireddyforpostgres(at)gmail(dot)com>, Dilip Kumar <dilipbalaut(at)gmail(dot)com>, "osumi(dot)takamichi(at)fujitsu(dot)com" <osumi(dot)takamichi(at)fujitsu(dot)com>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Allow logical replication to copy tables in binary format
Date: 2023-03-21 06:03:00
Message-ID: CAA4eK1JtAdWBVWO+6oYHN18LeY6w_t4n4rLwNhS4ifXk-PXk9Q@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Tue, Mar 21, 2023 at 7:03 AM Peter Smith <smithpb2250(at)gmail(dot)com> wrote:
>
>
> ======
> src/test/subscription/t/014_binary.pl
>
> 4.
> # -----------------------------------------------------
> # Test mismatched column types with/without binary mode
> # -----------------------------------------------------
>
> # Test syncing tables with mismatching column types
> $node_publisher->safe_psql(
> 'postgres', qq(
> CREATE TABLE public.test_mismatching_types (
> a bigint PRIMARY KEY
> );
> INSERT INTO public.test_mismatching_types (a)
> VALUES (1), (2);
> ));
>
> # Check the subscriber log from now on.
> $offset = -s $node_subscriber->logfile;
>
> # Ensure the subscription is enabled. disable_on_error is still true,
> # so the subscription can be disabled due to missing realtion until
> # the test_mismatching_types table is created.
> $node_subscriber->safe_psql(
> 'postgres', qq(
> CREATE TABLE public.test_mismatching_types (
> a int PRIMARY KEY
> );
> ALTER SUBSCRIPTION tsub ENABLE;
> ALTER SUBSCRIPTION tsub REFRESH PUBLICATION;
> ));
>
> ~~
>
> I found the "Ensure the subscription is enabled..." comment and the
> necessity for enabling the subscription to be confusing.
>
> Can't some complications all be eliminated just by creating the table
> on the subscribe side first?
>

Hmm, that would make this test inconsistent with other tests and
probably difficult to understand and extend. I don't like to say this
but I think introducing disable_on_error has introduced more
complexities in the patch due to the requirement of enabling
subscription again and again. I feel it would be better without using
disable_on_error option in these tests.

One minor point:
+ format can be faster than the text format, but it is less portable
+ across machine architectures and PostgreSQL versions.

As per email [1], it would be better to use <productname> tag here
with PostgreSQL.

[1] - https://www.postgresql.org/message-id/932629.1679322674%40sss.pgh.pa.us

--
With Regards,
Amit Kapila.

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Amit Kapila 2023-03-21 06:07:05 Re: Dropped and generated columns might cause wrong data on subs when REPLICA IDENTITY FULL
Previous Message John Naylor 2023-03-21 05:41:26 Re: [PoC] Improve dead tuple storage for lazy vacuum