Re: Support logical replication of DDLs

Поиск
Список
Период
Сортировка
От Masahiko Sawada
Тема Re: Support logical replication of DDLs
Дата
Msg-id CAD21AoBghkaJ4AzN0XE-4J6OCWH5JNQWops6eiwT0_FeGt4g+A@mail.gmail.com
обсуждение исходный текст
Ответ на Re: Support logical replication of DDLs  (Amit Kapila <amit.kapila16@gmail.com>)
Ответы Re: Support logical replication of DDLs  (Amit Kapila <amit.kapila16@gmail.com>)
Список pgsql-hackers
On Thu, Jun 23, 2022 at 7:00 PM Amit Kapila <amit.kapila16@gmail.com> wrote:
>
> On Wed, Jun 22, 2022 at 11:09 AM Masahiko Sawada <sawada.mshk@gmail.com> wrote:
> >
> > I've attached a WIP patch for adding regression tests for DDL deparse.
> > The patch can be applied on
> > v9-0001-Functions-to-deparse-DDL-commands.patch Hou recently
> > submitted[1]. The basic idea is to define the event trigger to deparse
> > DDLs, run the regression tests, load the deparsed DDLs to another
> > database cluster, dump both databases, and compare the dumps.
> >
>
> Thanks for working on this. It is a good start. I think this will be
> helpful to see the missing DDL support. Do we want to run this as part
> of every regression run? Won't it increase the regression time as this
> seems to run internally the regression tests twice?

Yes, It will increase the regression test time but we already do a
similar thing in 002_pg_upgrade.pl and 027_stream_regress.pl and it
seems to be worth adding to me.

>
> Do we need a different trigger to capture drop cases as there are
> separate deparsing routines for them, for example
> deparse_drop_table()?

Right, we need to capture drop cases by another trigger.

>
> > [2] deparsing "ALTER INDEX tbl_idx ALTER COLUMN 2 SET STATISTICS
> > 1000;" causes an assertion failure.
> >
>
> Sorry, it is not clear to me whether you are talking about some
> pre-existing bug or a bug in the proposed patch?

I meant there is a bug in the v9 DDL deparse patch.

Regards,


--
Masahiko Sawada
EDB:  https://www.enterprisedb.com/



В списке pgsql-hackers по дате отправления:

Предыдущее
От: "houzj.fnst@fujitsu.com"
Дата:
Сообщение: RE: Fix instability in subscription regression test
Следующее
От: John Naylor
Дата:
Сообщение: Re: NAMEDATALEN increase because of non-latin languages