On Tue, Apr 24, 2007 at 06:32:31PM -0400, [EMAIL PROTECTED] wrote:
> I installed postgresql 8.2.3 in a freebsd server, my client application is
> written in C++ builder + zeoslib and I haven't any problem until now, but now
> with 8.2.3 version I have many connection that remains in FIN_WAIT_2, any
Tom,
We were looking at the explain results and noticed that it was converting
the 'wynn%' into fname >= 'wynn' and fname < 'wyno'
So we modified our function to mimic this transformation, where we have
two variables that contain the 'wynn' and 'wyno':
Fname >= @var1 and fname < @var2
Listmail <[EMAIL PROTECTED]> writes:
>> Any idea why using a variable v_streetName instead of a string 'wynn%'
>> behaves differently?
> Yeah.
> 'wynn%' doesn't start with % so LIKE uses index.
> But Postgres doesn't know that you know that the contents of this
> variable never
tengo un indice sobre una columna date.
cuando hago una consulta ...
Select
id
from
ventas
WHERE
fecha = date('12-JAN-2007')
el planificador usa el indice
pero cuando la restriccion es WHERE fecha < date('12-JAN-2007')
se efectua un barrido secuencial.
ni e indice ni la tabl
Rich Shepard wrote:
> On Fri, 27 Apr 2007, Ron Johnson wrote:
>
>> Why put "active" data under /usr? Ever since user data was moved to
>> home,
>> /usr has only had system stuff in it. Or is that still a viable BSDism?
>
> Good question, Ron! But, in a multiuser system with many users of the
Any idea why using a variable v_streetName instead of a string 'wynn%'
behaves differently?
Yeah.
'wynn%' doesn't start with % so LIKE uses index.
But Postgres doesn't know that you know that the contents of this
variable never starts with '%'...
Thanks,
John
---
On Fri, 27 Apr 2007, Ron Johnson wrote:
Why put "active" data under /usr? Ever since user data was moved to home,
/usr has only had system stuff in it. Or is that still a viable BSDism?
Good question, Ron! But, in a multiuser system with many users of the
database, in whose /home directory
Schwenker, Stephen wrote:
> Hello,
>
> I've just compiled an instance of Postgresql 8.2.3 on a new linux box
> and have added some databases to it. I've noticed however that the
> autovacuum is not running. I have turned on the autovacuum,
> stats_start_collector, and stats_row_level and still
Marcelo de Moraes Serpa wrote:
> Hello,
>
> I have compiled PostgreSQL 8.3devel using Microsoft Visual C++ 2005 but
> when I tried to start it (either by using the initdb.exe or
> postgres.exe) I receive the following error message:
>
> Message title: " postgres.exe - Entry point not found"
> Mes
[EMAIL PROTECTED] ("John D. Burger") writes:
> I wrote:
>
>> I use a variant of The Tom Lane Solution previously pointed to,
>> your Plan 1 is very similar.
>
> Hmm, per that pointed-to post:
>
> http://archives.postgresql.org/pgsql-general/2003-05/msg00351.php
>
> I decided to run a periodic vac
Hello,
I have compiled PostgreSQL 8.3devel using Microsoft Visual C++ 2005 but when
I tried to start it (either by using the initdb.exe or postgres.exe) I
receive the following error message:
Message title: "postgres.exe - Entry point not found"
Message body: "It wasn't possible to find the entr
On Fri, 2007-04-27 at 11:09 -0700, RPK wrote:
> I am using pgsql 8.2.3 on windows XP. I noticed that dates are store in
> 'Date' columns of a table in "-mm-dd" format whereas I want to store it
> in "dd-MMM-" format. Can I change this?
First of all, you're talking about the input/output fo
In response to RPK <[EMAIL PROTECTED]>:
>
> I am using pgsql 8.2.3 on windows XP. I noticed that dates are store in
> 'Date' columns of a table in "-mm-dd" format whereas I want to store it
> in "dd-MMM-" format. Can I change this?
They're not stored in that format, they're just displayin
I am using pgsql 8.2.3 on windows XP. I noticed that dates are store in
'Date' columns of a table in "-mm-dd" format whereas I want to store it
in "dd-MMM-" format. Can I change this?
--
View this message in context:
http://www.nabble.com/Store-dates-in-database-in-different-format-tf365
Hello,
I need some help with a query inside a function. The table has 31M rows
in it (us streets), and a select statement takes 10ms when using an index
built for the query...
But, when the same query is in a function and uses a variable the index
isn't used and the query takes a few minutes.
I decided to run a periodic vacuum on my work queue. Lo and behold,
I get this:
ERROR: tuple concurrently updated
Which PG version is this, and do you have autovacuum enabled?
7.4.8, so no autovac, right?
In addition, all of my workers locked up, apparently indefinitely.
[ squint..
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 04/27/07 09:50, Rich Shepard wrote:
[snip]
>
> I see in postgresql.SlackBuild where the data directories are specified.
> I'll change that from /var/lib/pgsql/data to /usr/local/pgsql/data and
> rebuild the package.
(Explicitly noting that this
"John D. Burger" <[EMAIL PROTECTED]> writes:
> I decided to run a periodic vacuum on my work queue. Lo and behold,
> I get this:
>ERROR: tuple concurrently updated
Which PG version is this, and do you have autovacuum enabled? Awhile
back it was possible to get this error if autovac and a
I wrote:
I use a variant of The Tom Lane Solution previously pointed to,
your Plan 1 is very similar.
Hmm, per that pointed-to post:
http://archives.postgresql.org/pgsql-general/2003-05/msg00351.php
I decided to run a periodic vacuum on my work queue. Lo and behold,
I get this:
ERR
On Fri, 27 Apr 2007, Tom Lane wrote:
No, it hasn't. The PG code itself doesn't make any assumptions about the
location of $PGDATA AFAIR, but there are a large number of packagings that
assume specific places like /var/lib/pgsql/. If you've changed platforms
or changed between installing from sou
Marcelo de Moraes Serpa escribió:
> Hello!
>
> Today, I have finally managed to compile PostgreSQL 8.3devel using Visual
> C++ 2005 on Windows XP PRO. It was a tough process as I've never hacked such
> a big OS project like PGSQL. I did it becouse I needed to develop some C
> functions that will b
Rich Shepard <[EMAIL PROTECTED]> writes:
>Seems to me that the preferred filesystem for postgres has varied over the
> years from /var/lib/ to /usr/local/ and back again.
No, it hasn't. The PG code itself doesn't make any assumptions about
the location of $PGDATA AFAIR, but there are a large
Hello!
Today, I have finally managed to compile PostgreSQL 8.3devel using Visual
C++ 2005 on Windows XP PRO. It was a tough process as I've never hacked such
a big OS project like PGSQL. I did it becouse I needed to develop some C
functions that will be exported as functions to be called as Store
Seems to me that the preferred filesystem for postgres has varied over the
years from /var/lib/ to /usr/local/ and back again. When I upgraded from
7.4.3 to 8.1.4 the move was from /usr/local/ to /var/lib/. Now, as I prepare
to upgrade and migrate data from 8.1.4 to 8.2.4, I see that the docs al
Martijn van Oosterhout wrote:
> On Thu, Apr 26, 2007 at 12:13:13AM +0200, Listmail wrote:
> > VACUUM FULL is slow because it plays with the indexes...
> > CLUSTER is slow because it has to order the rows...
>
> And:
> VACUUM FULL has to seek/read/write all over the disk to get it's job
> d
Richard Huxton <[EMAIL PROTECTED]> writes:
>> SELECT '00:10:00'::TIME / '1:00:00'::TIME; -- Answer is:
>> 0.1667
>
> But it doesn't make any sense to divide one time by another, does it? Are you
> sure it's not intervals you want?
It "doesn't matter", I just used time because I
Jorge Godoy wrote:
Hi!
I am needing to convert from TIME type to floats and taking 1h as the
unit, so what I'd like is something that would allow me doing operations
like:
SELECT '00:10:00'::TIME / '1:00:00'::TIME; -- Answer is: 0.1667
But it doesn't make any sense to divi
On Thu, Apr 26, 2007 at 12:13:13AM +0200, Listmail wrote:
> VACUUM FULL is slow because it plays with the indexes...
> CLUSTER is slow because it has to order the rows...
And:
VACUUM FULL has to seek/read/write all over the disk to get it's job
done.
CLUSTER can scan through the table
Hi!
I am needing to convert from TIME type to floats and taking 1h as the
unit, so what I'd like is something that would allow me doing operations
like:
SELECT '00:10:00'::TIME / '1:00:00'::TIME; -- Answer is: 0.1667
SELECT '00:30:00'::TIME / '1:00:00'::TIME; -- Answer is: 0.
Hi,
I have some doubts regarding the settings and the access procedure of
warm standby servers:
- can autovacuum be safely enabled on the replicator?
- I'm using pg_standby (from cvs) that is generally working well as
expected (logs are copied with
scp); today I wanted to temporarily stop the
am Fri, dem 27.04.2007, um 8:18:55 +0100 mailte Peter Childs folgendes:
> >Sure, that's an alternative, but it seems a bit too much reinventing an
> >already existing wheel. I was amused to see the script had the ill-fated
> >Lee-Jackson-King day in it. Ideally, we'd want a Postgres table that
>
Felix Kater wrote:
Hi,
what is the binary representation of 'date' and 'numeric' data?
I am using PQexecParam (C binding libpq-8.0.12) with binary export
switched on. From the three examples in C code the types 'int4', 'bytea'
and strings are covered, however nothing about 'date', 'numeric'.
I
Hi,
what is the binary representation of 'date' and 'numeric' data?
I am using PQexecParam (C binding libpq-8.0.12) with binary export
switched on. From the three examples in C code the types 'int4', 'bytea'
and strings are covered, however nothing about 'date', 'numeric'.
If there is no documen
On 27.04.2007 03:09, CAJ CAJ wrote:
Forgive me if this has been discussed before (or if it sounds absurd)
Upgrading large postgres databases (100GB+) takes awfully long time when
doing dump/restore. I was wondering if this process can be optimized by
directly dumping to a new version of Pg datab
On 26/04/07, Greg Sabino Mullane <[EMAIL PROTECTED]> wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: RIPEMD160
>> However, you quickly run into the problem of holidays. While you
>> could construct a helper table listing all the holidays, ones that
>> don't fall on the same day every year (e.g
35 matches
Mail list logo
