Re: One source of constant annoyance identified

Hi!

I did as you suggested - and locale is indeed not set correctly, it
seems. Here's the output:

pg_control version number: 71
[...]
LC_COLLATE: C
LC_CTYPE: de_DE(at)euro

So LC_COLLATE wasn't set correctly; I dumped all databases, recompiled
(without multibyte- and recode-support this time, just to reduce
possible sources of failure), did initdb again and reimported the
dumpfile.

pg_control-output is now:
pg_control version number: 71
[...]
LC_COLLATE: de_DE(at)euro
LC_CTYPE: de_DE(at)euro

Configure-options were ./configure --prefix=/opt/pgsql/ --with-perl
--enable-odbc --enable-locale --enable-syslog

So that's most definitely fixed. It didn't do anything for me, though -
there are still several giant backends, as you can see by top-output

12:29pm up 1 day, 21:04, 3 users, load average: 1.44, 1.13, 1.18
85 processes: 81 sleeping, 4 running, 0 zombie, 0 stopped
CPU0 states: 87.3% user, 9.2% system, 0.0% nice, 3.1% idle
CPU1 states: 34.0% user, 6.0% system, 0.0% nice, 59.5% idle
CPU2 states: 47.3% user, 1.4% system, 0.0% nice, 50.5% idle
CPU3 states: 44.6% user, 7.3% system, 0.0% nice, 47.3% idle
Mem: 1029400K av, 1002272K used, 27128K free, 0K shrd, 5352K
buff
Swap: 2097136K av, 13160K used, 2083976K free 842792K
cached
PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND
9269 postgres 10 0 257M 257M 256M S 96.2 25.5 0:54 postmaster
9264 postgres 9 0 250M 250M 249M S 1.9 24.9 0:42 postmaster
9375 postgres 9 0 127M 127M 126M S 1.2 12.6 0:05 postmaster
9372 postgres 9 0 106M 106M 105M S 3.9 10.5 0:26 postmaster
9312 postgres 9 0 68900 67M 67616 S 1.8 6.6 0:02 postmaster
9379 postgres 9 0 41144 40M 39880 S 2.8 3.9 0:04 postmaster
9377 postgres 9 0 39276 38M 38104 S 0.0 3.8 0:14 postmaster
9380 postgres 9 0 34548 33M 33388 S 0.0 3.3 1:14 postmaster
9381 postgres 9 0 32140 31M 30760 S 0.0 3.1 0:26 postmaster
9373 postgres 9 0 24068 23M 23064 S 0.0 2.3 0:01 postmaster
9353 postgres 9 0 22224 21M 20332 S 7.8 2.1 0:37 postmaster
9371 postgres 9 0 22036 21M 21032 S 0.0 2.1 0:01 postmaster
9382 postgres 9 0 17840 17M 16652 S 0.0 1.7 0:01 postmaster
9268 postgres 12 0 16104 15M 15036 R 0.7 1.5 0:16 postmaster
9352 postgres 9 0 15584 15M 14076 S 0.0 1.5 0:19 postmaster
9351 postgres 19 0 14644 14M 13080 R 36.9 1.4 0:16 postmaster
9313 postgres 9 0 14068 13M 13036 S 5.1 1.3 0:14 postmaster
9374 postgres 9 0 13996 13M 12976 S 3.1 1.3 0:14 postmaster
9354 postgres 9 0 13116 12M 12088 S 1.3 1.2 0:03 postmaster
9376 postgres 9 0 12744 12M 11732 S 0.9 1.2 0:02 postmaster
9452 postgres 9 0 12356 12M 11208 S 0.0 1.1 0:10 postmaster
9370 postgres 9 0 12300 12M 11332 S 1.8 1.1 0:02 postmaster
9355 postgres 9 0 11768 11M 10756 S 0.0 1.1 0:00 postmaster
9524 postgres 9 0 10744 10M 9740 S 3.6 1.0 0:00 postmaster
9476 postgres 9 0 10020 9.8M 9044 S 1.9 0.9 0:00 postmaster
9602 postgres 18 0 9472 9468 8356 R 58.0 0.9 0:03 postmaster
9535 postgres 9 0 9368 9364 8216 S 0.7 0.9 0:00 postmaster

Any more ideas?

Regards,

Markus

> -----Ursprüngliche Nachricht-----
> Von: Tom Lane [mailto:tgl(at)sss(dot)pgh(dot)pa(dot)us]
> Gesendet: Freitag, 28. Juni 2002 22:42
> An: Markus Wollny
> Cc: Thomas Beutin; pgsql-general(at)postgresql(dot)org
> Betreff: Re: [GENERAL] One source of constant annoyance identified
>
>
> If you're not certain about this, compile
> contrib/pg_controldata and see
> what it says the locale values in your pg_control file are.
>
> regards, tom lane
>