Re: pg_dump issue : Cannot drop a non-existent(?) trigger - Mailing list pgsql-hackers

From Stephan Szabo
Subject Re: pg_dump issue : Cannot drop a non-existent(?) trigger
Date
Msg-id 20050325055728.A98363@megazone.bigpanda.com
Whole thread Raw
In response to pg_dump issue : Cannot drop a non-existent(?) trigger  (Devrim GUNDUZ <devrim@gunduz.org>)
Responses Re: pg_dump issue : Cannot drop a non-existent(?) trigger  (Tom Lane <tgl@sss.pgh.pa.us>)
Re: pg_dump issue : Cannot drop a non-existent(?) trigger  (Devrim GUNDUZ <devrim@gunduz.org>)
List pgsql-hackers
On Fri, 25 Mar 2005, Devrim GUNDUZ wrote:

> - --------------------------------------------------------------------------------------------------------------
>   PostgreSQL 8.0.1 on i686-redhat-linux-gnu, compiled by GCC gcc (GCC)
> 3.2.3 20030502 (Red Hat Linux 3.2.3-42)
> =======================================================================
>
> We wanted to take a full backup, so ran pg_dump,but got an error:
>
> =======================================================================
> $ pg_dump prod -U postgres > all.pgdump
> pg_dump: invalid argument string (firma_moduller_firma_fkey1) for trigger
> "RI_ConstraintTrigger_39053" on table "t_firma_moduller"
> =======================================================================
>
> Looking at the details, we saw the trigger there:
>
> =======================================================================
> prod=# \d t_firma_moduller
>                                                Table
> "public.t_firma_moduller"
>        Column      |            Type             |
> Modifiers
> -
------------------+-----------------------------+--------------------------------------------------------------------------
>   firma_no         | character varying(10)       | not null
>   modul_adi        | character varying(20)       | not null
>   last_update_date | timestamp without time zone |
>   last_update_user | character varying(45)       |
>   kod              | integer                     | not null default
> fn_get_seq_value('t_firma_moduller'::character varying)
> Indexes:
>      "t_firma_moduller_pkey" PRIMARY KEY, btree (firma_no, modul_adi)
>      "t_firma_moduller_kodu_ukey" UNIQUE, btree (kod)
> Triggers:
>      "RI_ConstraintTrigger_39053" AFTER INSERT OR UPDATE ON
> t_firma_moduller FROM t_firmalar NOT DEFERRABLE INITIALLY IMMEDIATE FOR
> EACH ROW EXECUTE PROCEDURE
> "RI_FKey_check_ins"('firma_moduller_firma_fkey1', 't_firma_moduller',
> 't_firmalar', 'UNSPECIFIED', 'firma_no', 'no')

Odd, why is this not showing up as a foreign key constraint?  Has this
database been upgraded (no matter through how many different upgrades)
from an old version (7.1 or earlier I think)?

> prod=# DROP TRIGGER RI_ConstraintTrigger_39053 ON t_firma_moduller;
> ERROR:  trigger "ri_constrainttrigger_39053" for table "t_firma_moduller"
> does not exist

You'd need to quote the trigger name because it's mixed case.

> Is this a bug or something? There should not be a trigger there.

Well, it really should be showing up as something like:
CONSTRAINT firma_moduller_firma_fkey1 FOREIGN KEY (firma_no)REFERENCES t_firmalar(no)



pgsql-hackers by date:

Previous
From: Palle Girgensohn
Date:
Subject: Re: Patch for collation using ICU
Next
From: Tom Lane
Date:
Subject: Re: Patch for collation using ICU