pg_dump/restore failure (dependency?) on BF serinus - Mailing list pgsql-hackers

From Andres Freund
Subject pg_dump/restore failure (dependency?) on BF serinus
Date
Msg-id l4joqljvbou26unhplkdj7m5olxn7rsy2loqlcta44nie5k3bs@zudmva7han7g
Whole thread Raw
Responses Re: pg_dump/restore failure (dependency?) on BF serinus
List pgsql-hackers
Hi,


https://buildfarm.postgresql.org/cgi-bin/show_stage_log.pl?nm=serinus&dt=2025-04-07%2017%3A45%3A28&stg=pg_upgrade-check
failed in a problem-indicating way in the new dump/restore test that Ashutosh
added.

> # Running: pg_restore --create -j2 -d postgres
/home/bf/bf-build/serinus/HEAD/pgsql.build/testrun/pg_upgrade/002_pg_upgrade/data/tmp_test_KTss/regression.dump
> pg_restore: error: could not execute query: ERROR:  there is no unique constraint matching given keys for referenced
table"pk"
 
> Command was: ALTER TABLE fkpart5.fk
>     ADD CONSTRAINT fk_a_fkey FOREIGN KEY (a) REFERENCES fkpart5.pk(a);
> 
> 
> pg_restore: warning: errors ignored on restore: 2

There are no other errors visible, neither in regress_log, nor in the server
log.

It's hard to tell without more logging, but the most likely explanation for
this seems that somehow the primary key on fkpart5.pk hasn't yet been
restored. However, looking at the pg_restore -v -l of the regression test
database locally, the relevant dependencies look sane on a first glance.

So I'm a bit confused how this could happen?

Greetings,

Andres Freund



pgsql-hackers by date:

Previous
From: Yugo NAGATA
Date:
Subject: Re: Extend ALTER DEFAULT PRIVILEGES for large objects
Next
From: Tom Lane
Date:
Subject: Re: pg_dump/restore failure (dependency?) on BF serinus