Re: Karsten Hilbert 2017-10-07
<20171007155734.jf2q4cz3ghcdn...@hermes.hilbert.loc>
> Digging deeper: pg_upgradecluster does
> $encoding comes from get_db_encoding
> which also uses psql, which needs -X, too, in PgCommon.pm
Ah right - I already fixed that in the commit I pushed
On Sat, Oct 07, 2017 at 05:46:07PM +0200, Karsten Hilbert wrote:
> > Can you check if the upgrade works properly if you remove the
> > "unless"?
> >
> > chomp $ctype;
> > chomp $collate;
> > print STDERR "$ctype / $collate\n";
> > return ($ctype, $collate);
>
> Unfortunately
Re: Karsten Hilbert 2017-10-07
<20171007154607.34udpgu2jxwsr...@hermes.hilbert.loc>
> > chomp $ctype;
> > chomp $collate;
> > print STDERR "$ctype / $collate\n";
> > return ($ctype, $collate);
>
> Unfortunately not:
>
> root@hermes:~/bin# pg_upgradecluster 9.6 main
>
Re: Karsten Hilbert 2017-10-07
<20171007154139.bpvnx7477qcyh...@hermes.hilbert.loc>
> condition: $? which is 0 which is "FALSE"
0 is a program exit code where 0 is true and everything else is false.
So yes, this is inverted logic.
Christoph
On Sat, Oct 07, 2017 at 04:40:24PM +0200, Christoph Berg wrote:
> Can you check if the upgrade works properly if you remove the
> "unless"?
>
> chomp $ctype;
> chomp $collate;
> print STDERR "$ctype / $collate\n";
> return ($ctype, $collate);
Unfortunately not:
On Sat, Oct 07, 2017 at 05:34:15PM +0200, Christoph Berg wrote:
> Re: Karsten Hilbert 2017-10-07
> <20171007152609.i4m5zc6yuol4j...@hermes.hilbert.loc>
return ($ctype, $collate) unless $?;
> >
> > Actually, isn't this inverted logic ?
> >
> > It should return(...) unless $? actually
Re: Karsten Hilbert 2017-10-07
<20171007152609.i4m5zc6yuol4j...@hermes.hilbert.loc>
> > > return ($ctype, $collate) unless $?;
>
> Actually, isn't this inverted logic ?
>
> It should return(...) unless $? actually is NOT zero ?
The extracted values should be returned if the last psql call
On Sat, Oct 07, 2017 at 04:40:24PM +0200, Christoph Berg wrote:
> > return ($ctype, $collate) unless $?;
Actually, isn't this inverted logic ?
It should return(...) unless $? actually is NOT zero ?
PSQL(1)
PostgreSQL
On Sat, Oct 07, 2017 at 04:40:24PM +0200, Christoph Berg wrote:
> Re: Karsten Hilbert 2017-10-07
> <20171007132123.7eqyzz7455f5x...@hermes.hilbert.loc>
> > root@hermes:/usr/share/perl5# pg_upgradecluster 9.6 main
> > de_DE.UTF-8 / de_DE.UTF-8
> > Error: could not get cluster locales
Re: Karsten Hilbert 2017-10-07
<20171007132123.7eqyzz7455f5x...@hermes.hilbert.loc>
> root@hermes:/usr/share/perl5# pg_upgradecluster 9.6 main
> de_DE.UTF-8 / de_DE.UTF-8
> Error: could not get cluster locales
Hmm. At least the uninitialized value warning is gone now.
> I also
On Sat, Oct 07, 2017 at 02:11:14PM +0200, Christoph Berg wrote:
> > postgres@hermes:~$ psql -d template1
> > Ausgabeformat ist »wrapped«.
>
> PgCommon.pm doesn't use 'psql -X' so your (or postgres') .psqlrc might
> be getting in the way. Could you try the upgrade again after moving
>
Re: Karsten Hilbert 2017-10-07
<20171007104018.ax3cv7dwqg4s3...@hermes.hilbert.loc>
> Thanks for your response !
Thanks!
> postgres@hermes:~$ psql -d template1
> Ausgabeformat ist »wrapped«.
PgCommon.pm doesn't use 'psql -X' so your (or postgres') .psqlrc might
be getting in the
Thanks for your response !
> could you provide the output of 'psql -l' for the old cluster?
Sure:
postgres@hermes:~$ psql -l
Ausgabeformat ist »wrapped«.
Liste der Datenbanken
Name| Eigentümer | Kodierung | Sortierfolge |
Control: tags -1 moreinfo
Re: Karsten Hilbert 2017-10-07
<150737030347.16226.787705667158988632.report...@hermes.hilbert.loc>
> Use of uninitialized value $ctype in scalar chomp at
> /usr/share/perl5/PgCommon.pm line 947.
> Use of uninitialized value $collate in scalar chomp at
>
Package: postgresql-common
Version: 186
Severity: important
Hello,
when upgrading
Ver Cluster Port Status OwnerData directory Log file
9.6 main5432 online postgres /var/lib/postgresql/9.6/main
/var/log/postgresql/postgresql-9.6-main.log
like so:
15 matches
Mail list logo
