Re: Add pg_walinspect function with block info columns - Mailing list pgsql-hackers

From Bharath Rupireddy
Subject Re: Add pg_walinspect function with block info columns
Date
Msg-id CALj2ACXhfSAtDZA+kHhb+ODe0qyPLv1zvYBtcnhKwrLSe_=2aA@mail.gmail.com
Whole thread Raw
In response to Re: Add pg_walinspect function with block info columns  (Kyotaro Horiguchi <horikyota.ntt@gmail.com>)
Responses Re: Add pg_walinspect function with block info columns
List pgsql-hackers
On Thu, Mar 9, 2023 at 7:34 AM Kyotaro Horiguchi
<horikyota.ntt@gmail.com> wrote:
>
> > In short, my choice would still be simplicity here with v3, I guess.
>
> FWIW, I slightly prefer v3 for the reason I mentioned above.

Hm, then, +1 for v3.

FWIW, I quickly tried to hit that case where a single WAL record has
max_block_id = XLR_MAX_BLOCK_ID with both FPIs and block data, but I
couldn't. I could generate WAL records with 45K FPIs, 10mn block data
and the total palloc'd length in the block_id for loop has not crossed
8K.

--
Bharath Rupireddy
PostgreSQL Contributors Team
RDS Open Source Databases
Amazon Web Services: https://aws.amazon.com



pgsql-hackers by date:

Previous
From: Michael Paquier
Date:
Subject: Re: meson vs make: missing/inconsistent ENV
Next
From: vignesh C
Date:
Subject: Re: [PATCH] Use indexes on the subscriber when REPLICA IDENTITY is full on the publisher