[Bug] pg_relation_size may get wrong result.

[shmiwy]

Cloudberry Database version

No response

What happened

SET gp_create_table_random_default_distribution=off;
create or replace function junkloop(rel text, numiter int) returns int as $$
declare
  sql text;
begin
  for i in 0..numiter loop
    sql := $sql$ insert into $sql$ || rel || $sql$ select 1, $sql$ || i::text || $sql$, repeat('x', 1000) $sql$;
    execute sql;
  end loop;
  return numiter;
end;
$$ language plpgsql;

drop table if exists vfao;
create table vfao (a, b, c) with (appendonly=true, orientation=column) as
select 1, i, repeat('x', 1000) from generate_series(1, 100)i distributed by (a);
create index ivfao on vfao(b, c);

select junkloop('vfao', 300);
select pg_relation_size((select segrelid from pg_appendonly where relid = 'vfao'::regclass)) from gp_dist_random('gp_id') where gp_segment_id = 1;

the result for now is

 pg_relation_size
------------------
           163840

It looks like this value is too large.

What you think should happen instead

No response

How to reproduce

rerun the query.

Operating System

centos7

Anything else

No response

Are you willing to submit PR?

• Yes, I am willing to submit a PR!

Code of Conduct

• I agree to follow this project's Code of Conduct.

[github-actions]

Hey, @shmiwy welcome!🎊 Thanks for taking the time to point this out.🙌