Re: Add a new pg_walinspect function to extract FPIs from WAL records

On Thu, Jan 5, 2023 at 6:51 PM Bharath Rupireddy
<bharath(dot)rupireddyforpostgres(at)gmail(dot)com> wrote:
>
> > I'm also wondering if it would make sense to extend the test coverage of it (and pg_waldump) to "validate" that both
> > extracted images are the same and matches the one modified right after the checkpoint.
> >
> > What do you think? (could be done later in another patch though).
>
> I think pageinspect can be used here. We can fetch the raw page from
> the table after the checkpoint and raw FPI from the WAL record logged
> as part of the update. I've tried to do so [1], but I see a slight
> difference in the raw output. The expectation is that they both be the
> same. It might be that the update operation logs the FPI with some
> more info set (prune_xid). I'll try to see why it is so.
>
> I'm attaching the v2 patch for further review.
>
> [1]
> SELECT * FROM page_header(:'page_from_table');
> lsn | checksum | flags | lower | upper | special | pagesize |
> version | prune_xid
> -----------+----------+-------+-------+-------+---------+----------+---------+-----------
> 0/1891D78 | 0 | 0 | 40 | 8064 | 8192 | 8192 |
> 4 | 0
> (1 row)
>
> SELECT * FROM page_header(:'page_from_wal');
> lsn | checksum | flags | lower | upper | special | pagesize |
> version | prune_xid
> -----------+----------+-------+-------+-------+---------+----------+---------+-----------
> 0/1891D78 | 0 | 0 | 44 | 8032 | 8192 | 8192 |
> 4 | 735
> (1 row)

Ugh, v2 patch missed the new file added, I'm attaching v3 patch for
further review. Sorry for the noise.

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