Re: [PoC] Improve dead tuple storage for lazy vacuum - Mailing list pgsql-hackers

From John Naylor
Subject Re: [PoC] Improve dead tuple storage for lazy vacuum
Date
Msg-id CAFBsxsHV9bqP3MKeHY4fcm-4pQF-S22RtV2tx=M3LhaCXzBbxA@mail.gmail.com
Whole thread Raw
In response to Re: [PoC] Improve dead tuple storage for lazy vacuum  (Masahiko Sawada <sawada.mshk@gmail.com>)
Responses Re: [PoC] Improve dead tuple storage for lazy vacuum  (Masahiko Sawada <sawada.mshk@gmail.com>)
List pgsql-hackers
On Sun, Aug 27, 2023 at 7:53 PM Masahiko Sawada <sawada.mshk@gmail.com> wrote:
>
> I've updated the regression tests for tidstore so that it uses SQL
> functions to add blocks/offsets and dump its contents. The new test
> covers the same test coverages but it's executed using SQL functions
> instead of executing all tests in one SQL function.

This is much nicer and more flexible, thanks! A few questions/comments:

tidstore_dump_tids() returns a string -- is it difficult to turn this into a SRF, or is it just a bit more work?

The lookup test seems fine for now. The output would look nicer with an "order by tid".

I think we could have the SQL function tidstore_create() take a boolean for shared memory. That would allow ad-hoc testing without a recompile, if I'm not mistaken.

+SELECT tidstore_set_block_offsets(blk, array_agg(offsets.off)::int2[])
+  FROM blocks, offsets
+  GROUP BY blk;
+ tidstore_set_block_offsets
+----------------------------
+
+
+
+
+
+(5 rows)

Calling a void function multiple times leads to vertical whitespace, which looks a bit strange and may look better with some output, even if irrelevant:

-SELECT tidstore_set_block_offsets(blk, array_agg(offsets.off)::int2[])
+SELECT row_number() over(order by blk), tidstore_set_block_offsets(blk, array_agg(offsets.off)::int2[])

 row_number | tidstore_set_block_offsets
------------+----------------------------
          1 |
          2 |
          3 |
          4 |
          5 |
(5 rows)

--
John Naylor
EDB: http://www.enterprisedb.com

pgsql-hackers by date:

Previous
From: Kyotaro Horiguchi
Date:
Subject: Strange presentaion related to inheritance in \d+
Next
From: Peter Smith
Date:
Subject: Re: [PoC] pg_upgrade: allow to upgrade publisher node