Hi,
A DSS developer from my company, Julien Roze, reported me an error I cannot
explained. Is it a new behavior or a bug ?
Original query is much more complicated but here is a simplified test case with
postgresql 14 and 15 beta 2 on Debian 11, packages from pgdg :
Ver Cluster Port Status Owner Data directory Log file
14 main 5432 online postgres /var/lib/postgresql/14/main
/var/log/postgresql/postgresql-14-main.log
15 main 5433 online postgres /var/lib/postgresql/15/main
/var/log/postgresql/postgresql-15-main.log
psql -p 5432
select version();
version
-----------------------------------------------------------------------------------------------------------------------------
PostgreSQL 14.4 (Debian 14.4-1.pgdg110+1) on x86_64-pc-linux-gnu, compiled by
gcc (Debian 10.2.1-6) 10.2.1 20210110, 64-bit
(1 ligne)
with fakedata as (
select 'hello' word
union all
select 'world' word
)
select *
from (
select word, count(*) over (partition by word) nb from fakedata
) t where nb = 1;
word | nb
-------+----
hello | 1
world | 1
(2 lignes)
with fakedata as (
select 'hello' word
union all
select 'world' word
)
select *
from (
select word, count(*) nb from fakedata group by word
) t where nb = 1;
word | nb
-------+----
hello | 1
world | 1
(2 lignes)
psql -p 5433
select version();
version
------------------------------------------------------------------------------------------------------------------------------------
PostgreSQL 15beta2 (Debian 15~beta2-1.pgdg110+1) on x86_64-pc-linux-gnu,
compiled by gcc (Debian 10.2.1-6) 10.2.1 20210110, 64-bit
(1 ligne)
with fakedata as (
select 'hello' word
union all
select 'world' word
)
select *
from (
select word, count(*) over (partition by word) nb from fakedata
) t where nb = 1;
ERREUR: cache lookup failed for function 0
with fakedata as (
select 'hello' word
union all
select 'world' word
)
select *
from (
select word, count(*) nb from fakedata group by word
) t where nb = 1;
word | nb
-------+----
hello | 1
world | 1
(2 lignes)
Best regards,
Phil
