Re: aaagh... postgres is segfaulting? -- fix - Mailing list pgsql-general

From Joshua Moore-Oliva
Subject Re: aaagh... postgres is segfaulting? -- fix
Date
Msg-id 200304091021.49706.josh@chatgris.com
Whole thread Raw
In response to Re: aaagh... postgres is segfaulting? -- fix  (Richard Huxton <dev@archonet.com>)
List pgsql-general

On April 9, 2003 09:35 am, Richard Huxton wrote:
> On Wednesday 09 Apr 2003 12:25 pm, Joshua Moore-Oliva wrote:
> > I needed to apply this...  for some reason when I do timetamp related
> > queries without executing that a segfault occurs.
> >
> > UPDATE pg_proc SET prorettype = 1114 WHERE prosrc = 'timestamptz_izone';
>
> Hmm - mine reads 1186 (interval). Two questions:
>
> 1. Did you notice what value it had before?

Yea it had an interval before.

> 2. Have you made the same change in template0/1 (and what values did they
> have)?

no I didn't touch template 0 or 1, but it just dawned on me that doing so may
save a step for me having to run that command with every database I install.

>
> Actually, a third question: how did you ever figure that out?

A whle ago I had this problem..

SELECT now() AT TIME ZONE interval '-5 hours'

returns type interval not timestamp with//out time zone.

A prety knowledgable guy Tom Lane suggested the above fix.

When I was getting a segfault on a query similar to that it dawned on me that
i may need to run that command again since I had moved the database to a new
server, and presto it worked :)

Josh.


pgsql-general by date:

Previous
From: "enediel"
Date:
Subject: multiple fields index
Next
From: Stephan Szabo
Date:
Subject: Re: altering table