Re: [UDD] Role of different upstream status tables (upstream versus dehs)

[Andreas Tille <andreas@an3as.eu>]

On Wed, May 27, 2015 at 06:41:34PM +0200, Lucas Nussbaum wrote:
> > 
> > I'd consider the implementation of status as ENUM more performant than
> > a plain text.  Am I missing something?
> 
> status is returned by uscan. I did not want to write in stone the list
> of possible statuses returned by uscan.

Well, there are things written in stone that are more in flux than these
but I have no interest to discuss this here. 

What I'm a bit more concerned about is that the UDD dump has problems
to import exactly this table.  I just downloaded a fresh dump and the
log says:

FEHLER:  ungültige Byte-Sequenz für Kodierung „UTF8“: 0x99
CONTEXT:  COPY upstream, Zeile 235
FEHLER:  ungültige Byte-Sequenz für Kodierung „UTF8“: 0xf6 0x6c 0x6c 0xe4
CONTEXT:  COPY upstream_status, Zeile 8680

Sorry for German locale - may be this is even the source of the problem???
It somehow says "invalid byte sequence for UTF encoding".  I would
***really*** love to know how to revert the German postresql locale. :-(

Anyway I tried to reimport the table but got:



$ ./update-and-run.sh upstream
/usr/lib/ruby/vendor_ruby/dbd/pg/database.rb:82:in `rescue in initialize': FATAL:  Peer-Authentifizierung für Benutzer „udd“ fehlgeschlagen (DBI::OperationalError)
        from /usr/lib/ruby/vendor_ruby/dbd/pg/database.rb:41:in `initialize'
        from /usr/lib/ruby/vendor_ruby/dbd/Pg.rb:150:in `new'
        from /usr/lib/ruby/vendor_ruby/dbd/Pg.rb:150:in `connect'
        from /usr/lib/ruby/vendor_ruby/dbi/handles/driver.rb:33:in `connect'
        from /usr/lib/ruby/vendor_ruby/dbi.rb:137:in `connect'
        from /srv/udd.debian.org/udd/scripts/update-upstream-status:224:in `<main>'


It says:

   Peer-authentication for user 'udd' failed

What would be the correct pg_hba.conf line to enable ruby injecting data?

Thanks for any guidance

       Andreas.


-- 
http://fam-tille.de