Re: [ADMIN] pg_upgrade from 9.1.3 to 9.2 failed

Bruce Momjian <bruce@momjian.us>

From: Bruce Momjian <bruce@momjian.us>
To: Peter Eisentraut <peter_e@gmx.net>
Cc: Rural Hunter <ruralhunter@gmail.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>, Tom Lane <tgl@sss.pgh.pa.us>
Date: 2012-09-24T13:08:20Z
Lists: pgsql-hackers
On Mon, Sep 24, 2012 at 09:06:04AM -0400, Peter Eisentraut wrote:
> On 9/24/12 8:55 AM, Bruce Momjian wrote:
> > I can confirm that pg_upgrade does case-insensitive comparisons of
> > encoding/locale names:
> > 
> > 	static void
> > 	check_locale_and_encoding(ControlData *oldctrl,
> > 	                          ControlData *newctrl)
> > 	{
> > 	    /* These are often defined with inconsistent case, so use pg_strcasecmp(). */
> > 	    if (pg_strcasecmp(oldctrl->lc_collate, newctrl->lc_collate) != 0)
> > 	        pg_log(PG_FATAL,
> > 	               "old and new cluster lc_collate values do not match\n");
> > 	    if (pg_strcasecmp(oldctrl->lc_ctype, newctrl->lc_ctype) != 0)
> > 	        pg_log(PG_FATAL,
> > 	               "old and new cluster lc_ctype values do not match\n");
> 
> I seem to recall that at some point in the distant past, somehow some
> Linux distributions changed the canonical spelling of locale names from
> xx_YY.UTF-8 to xx_YY.utf8.  So if people are upgrading old PostgreSQL
> instances that use the old spelling, pg_upgrade will probably fail.  A
> fix might be to take the locale name you find in pg_control and run it
> through setlocale() to get the new canonical name.

Or we could just remove dashes from the name before comparisons.

-- 
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com

  + It's impossible for everything to be true. +


Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →
  1. Replace empty locale name with implied value in CREATE DATABASE and initdb.