Date: Wed, 23 Nov 2022 09:03:05 +0000 (UTC)
From: Boboc Cristi <bobocc@yahoo.com>
Reply-To: Boboc Cristi <bobocc@yahoo.com>
To: PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>,
	Peter Smith <smithpb2250@gmail.com>
Message-ID: <89363341.1728627.1669194185204@mail.yahoo.com>
In-Reply-To: 
 <CAHut+Pui7LCgfzkjcqvdcPy+V9G8CRb6X5hL+MB7QqXFBhQMzg@mail.gmail.com>
References: <166870632180.632.14582512493527657072@wrigleys.postgresql.org>
 <CAHut+Pui7LCgfzkjcqvdcPy+V9G8CRb6X5hL+MB7QqXFBhQMzg@mail.gmail.com>
Subject: Re: Logical replication missing information
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----=_Part_1728626_936620228.1669194185201"
Content-Length: 19741
Archived-At: 
 <https://www.postgresql.org/message-id/89363341.1728627.1669194185204%40mail.yahoo.com>
Precedence: bulk

------=_Part_1728626_936620228.1669194185201
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hello!Thank you very much, the information you just sent to me is very, ver=
y valuable!
Best regards, Cristi Boboc=20

    On Wednesday, November 23, 2022 at 04:45:09 AM GMT+2, Peter Smith <smit=
hpb2250@gmail.com> wrote: =20
=20
 On Fri, Nov 18, 2022 at 4:50 AM PG Doc comments form
<noreply@postgresql.org> wrote:
>
> The following documentation comment has been logged on the website:
>
> Page: https://www.postgresql.org/docs/15/logical-replication-row-filter.h=
tml
> Description:

Hi,

FYI - I have forwarded this post to the hacker's list, where I think
it will receive more attention.

I am not sure why that (above) page was cited -- the section "31.3 Row
Filters" is specifically about row filtering, whereas the items you
reported seem unrelated to row filters, but are generic for all
Logical Replication.

>
> There are several things missing here and some of them I found to be high=
ly
> important:
> 1. How can I find why a logical replication failed. Currently I only can =
see
> it "does nothing" in pg_stat_subscriptions.

There should be logs reporting any replication conflicts etc. See [1]
for example logs. See also the answer for #2 below.

> 2. In case of copying the existing data: how can I find which tables or
> partitions were processed and which are on the processing queue (while
> monitoring I have observed no specific order or rule).

There is no predictable processing queue or order - The initial
tablesyncs might be happening in multiple asynchronous processes
according to the GUC max_sync_workers_per_subscription [2].

Below I show examples of replicating two tables (tab1 and tab2).

~~