[BUGS] Problems with UTF8 in PSQL

2006-07-12 Thread rolf
Hallo,

I don't know if the problem has been reported yet, but it seems to have more 
implications.

- Prerequisite:
  o PostgreSQL 8.1.4 on i486-pc-linux-gnu, compiled by GCC cc (GCC) 4.0.4
 20060507 (prerelease) (Debian 4.0.3-3)
  o and psql
  o A database which was created with -E UTF8
- Description:
  If i log in into that database (client encoding is also in UTF8) a lot
  of  "\"-commands are not accepted and the error message says
  "invalid byte sequence for UTF8"

- Original output:
  test=> \d
  FEHLER:  ungültige Byte-Sequenz für Kodierung »UTF8«: 0xfc6d6572220a
  test=> \q

  Same error message is in the log file

-Implications?
  There are also problems in this field, if I try to add a constraint for a
  primary key from inside of another programming language with the same
  error message! So it could be a problem from PostqreSQL not from psql.

HTH to find the problem
Bye
rolf

---(end of broadcast)---
TIP 6: explain analyze is your friend


[BUGS] BUG #2524: Error while using pg_restore command

2006-07-12 Thread Sudarshan

The following bug has been logged online:

Bug reference:  2524
Logged by:  Sudarshan
Email address:  [EMAIL PROTECTED]
PostgreSQL version: 8.0
Operating system:   Windoes-XP
Description:Error while using pg_restore command
Details: 

When i run the pg_restore command

"pg_restore -d db1 -v -h localhost -p 5432 -U postgres db2.backup"

I get the following error

pg_restore: [archiver (db)] connection to database "db1" failed: FATAL: 
password authentication failed for user "postgres"

I checked and rechecked the password and it seems correct. I also saw so
many threads on "pg_restore" but could not find suitable answer for this.

Ealiest help greatly appreciated.

Cheers
Sudarshan

---(end of broadcast)---
TIP 4: Have you searched our list archives?

   http://archives.postgresql.org


Re: [BUGS] BUG #2524: Error while using pg_restore command

2006-07-12 Thread Magnus Hagander
> The following bug has been logged online:
> 
> Bug reference:  2524
> Logged by:  Sudarshan
> Email address:  [EMAIL PROTECTED]
> PostgreSQL version: 8.0
> Operating system:   Windoes-XP
> Description:Error while using pg_restore command
> Details: 
> 
> When i run the pg_restore command
> 
> "pg_restore -d db1 -v -h localhost -p 5432 -U postgres db2.backup"
> 
> I get the following error
> 
> pg_restore: [archiver (db)] connection to database "db1" 
> failed: FATAL: 
> password authentication failed for user "postgres"
> 
> I checked and rechecked the password and it seems correct. I 
> also saw so many threads on "pg_restore" but could not find 
> suitable answer for this.
> 
> Ealiest help greatly appreciated.

This looks like a bug that has been fixed in the latest versions in both
the 8.0 and the 8.1 series. Upgrade to 8.0.8 or 8.1.4 and see if that
fixes it.

//Magnus

---(end of broadcast)---
TIP 3: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faq


Re: [BUGS] Problems with UTF8 in PSQL

2006-07-12 Thread REISS Thomas DSIC DESP

Hello,

When I experience the problem, the current Linux locale on my server is 
[EMAIL PROTECTED] Psql works fine if I change the current locale (by setting 
the LANG shell variable) to [EMAIL PROTECTED]


I would be glad if psql may work fine without changing the LANG variable 
in the current session.


Regards,
Thomas


rolf a écrit :

Hallo,

I don't know if the problem has been reported yet, but it seems to have more 
implications.


- Prerequisite:
  o PostgreSQL 8.1.4 on i486-pc-linux-gnu, compiled by GCC cc (GCC) 4.0.4
 20060507 (prerelease) (Debian 4.0.3-3)
  o and psql
  o A database which was created with -E UTF8
- Description:
  If i log in into that database (client encoding is also in UTF8) a lot
  of  "\"-commands are not accepted and the error message says
  "invalid byte sequence for UTF8"

- Original output:
  test=> \d
  FEHLER:  ungültige Byte-Sequenz für Kodierung »UTF8«: 0xfc6d6572220a
  test=> \q

  Same error message is in the log file

-Implications?
  There are also problems in this field, if I try to add a constraint for a
  primary key from inside of another programming language with the same
  error message! So it could be a problem from PostqreSQL not from psql.

HTH to find the problem
Bye
rolf

---(end of broadcast)---
TIP 6: explain analyze is your friend
  


--

Thomas REISS
Ministère de l'Intérieur et de l'Aménagement du Territoire
SG/DSIC/SCSIC/DESP/BSTN
Tél. : 01-61-44-25-90
E-mail: [EMAIL PROTECTED] 



---(end of broadcast)---
TIP 2: Don't 'kill -9' the postmaster


[BUGS] BUG #2525: "No buffer space available" error running pg_dump on win2k for large bytea values

2006-07-12 Thread Steve McWilliams

The following bug has been logged online:

Bug reference:  2525
Logged by:  Steve McWilliams
Email address:  [EMAIL PROTECTED]
PostgreSQL version: 8.1.4
Operating system:   Windows 2000
Description:"No buffer space available"  error running pg_dump on
win2k for large bytea values
Details: 

We are having a problem with pg_dump that seems to be limited to windows
2000 platforms when we try to dump a db containing large bytea values.  When
I run pg_dump (options "-b -Fc -Z0") to dump the contents of a test database
that isolates this problem I get the following error printed to the
terminal:

pg_dump: could not receive data from server: No buffer space available
(0x2747/10055)
pg_dump: SQL command to dump contents of table 'my_table' failed:
PQendcopy() failed.
pg_dump: Error message from server: could not receive data from server: No
buffer space available (0x2747/10055)
pg_dump: The command was: COPY public.my_table (my_id, my_array) TO stdout;

The postmaster log file shows the following corresponding entry:

LOG:  could not receive data from client: No connection could be made
because the target machine actively refused it.

The database I am backing up here contains a single table.  The table
contains 2 columns, an integer id column and a bytea column.  The table
contains a single row with a bytea value containing 25.7 mb of binary data. 
If necessary I can send a dump of the db in question containing its schema
and data (generated on linux).

I am running postgres 8.1.4 on windows 2000 v5.00.2195 (srevice pack 4) on a
pentium 4 2.6 ghz dedicated machine with 784 mb physical memory and 1.5 gb
total virtual memory.  The same machine is running both the postgres server
and the pg_dump client, with nothing much else going on concurrently.  I do
not see the OS running out of resources during the pg_dump attempt (watching
via TaskManager).

We are not passing any options to postmaster at launch time other than the
"-i" option to tell it to use tcp sockets, and we are using the default
values in postgresql.conf.

Note we are not able to reproduce this on windows xp pro or on windows 2003,
or on linux.  On windows 2k however it seems to always happen.

A co-worker here reported this problem to pgsql-general several months ago. 
The relevent links are:

http://archives.postgresql.org/pgsql-general/2006-01/msg00328.php
http://archives.postgresql.org/pgsql-general/2006-01/msg00952.php
http://archives.postgresql.org/pgsql-general/2006-01/msg00927.php

At the time my co-worker thought he had worked around it by adding a sleep
to the pg_dump read loop.  This may have helped at the time (at the expense
of making dumps take much much longer) I'm not sure, however recently the
problem has returned so we need to find a better solution.

As pointed out in one of the responses in the link above, the underlying
problem seems to be that windows is returning a WSAENOBUFS error at some
point during the transfer of data over the socket involved.  The microsoft
link referred to is:

http://support.microsoft.com/kb/q201213

I'm not sure what the socket send/receive buffer sizes being used by
postgres are.  I tried hacking the postgres source where the sockets are
created (both server and client sockets) to set their options to make sure
they use buffer sizes less than 64k as suggested by microsoft, however that
did not affect the problem.  

The tweakable registry entries referred to I am guessing are the same ones
mentioned in another link I found when googling this problem:

http://wiki.noreply.org/noreply/TheOnionRouter/WindowsBufferProblems

I have not tried these registry tweaks yet, but from the description I don't
think they would help because they pertain to WSAENOBUFS problems that occur
when many sockets (5000+) are being used by the application, which is not
the case here.

---(end of broadcast)---
TIP 2: Don't 'kill -9' the postmaster


[BUGS] BUG #2526: records lost

2006-07-12 Thread wee goh

The following bug has been logged online:

Bug reference:  2526
Logged by:  wee goh
Email address:  [EMAIL PROTECTED]
PostgreSQL version: 8.0.3-2
Operating system:   linux
Description:records lost
Details: 

i discovered records in my postgresql database missing every 6 - 8 months. i
supsect those not updated or amended are the ones that are gone, still
wondering the real reason behind ? i doubt is lack of vaccum as my
transactions are minnimal, inserting about 200 records each month and
updating about 1000 records each month.

this happens with previous versions of postgresql as well.

---(end of broadcast)---
TIP 2: Don't 'kill -9' the postmaster


Re: [BUGS] bug report: recrods missing periodically

2006-07-12 Thread WP Goh
 i discovered records in my postgresql database missing every 6 - 8 months. i supsect those not updated or amended are the ones that are gone, still wondering the real reason behind ? i doubt is lack of vaccum as my transactions are minnimal, inserting about 200 records each month and updating about 1000 records each month.this happens with previous versions of postgresql as well.Thu, 13 Jul 2006 6:00:43  (Singapore) 
		Do you Yahoo!? Everyone is raving about the  all-new Yahoo! Mail Beta.

Re: [BUGS] BUG #2526: records lost

2006-07-12 Thread Michael Fuhr
On Wed, Jul 12, 2006 at 08:58:43PM +, wee goh wrote:
> i discovered records in my postgresql database missing every 6 - 8 months. i
> supsect those not updated or amended are the ones that are gone, still
> wondering the real reason behind ? i doubt is lack of vaccum as my
> transactions are minnimal, inserting about 200 records each month and
> updating about 1000 records each month.

How much querying are you doing?  Are you vacuuming at all?  What's
the output of the following command?

SELECT datname, age(datvacuumxid) AS vage, age(datfrozenxid) AS fage
FROM pg_database;

Do the server logs contain any unusual messages?  Have you had any
hardware problems or system crashes?

-- 
Michael Fuhr

---(end of broadcast)---
TIP 4: Have you searched our list archives?

   http://archives.postgresql.org