Jamie Deppeler wrote:
Hi,
having a small problem in a trigger i want to automatically update a
date field but when i use for example
update table
set field = current_date
where table.pk = new.pk
If "table" is the table with the trigger on it, you'll end up with an
endless loop, since the trigger
Postgres Learner wrote:
Hi all!
I recently started using pg_autovacuum instead of scheduling vacuum
analyzes in a system that needs to be taken to production soon.
However, I have noticed something funny that happens while using this.
sometimes, some database operations take an unreasonably long ti
Pg 7.4.5
Curious: Why would a DELETE FROM tablename work while a pg_dump is
occurring but a TRUNCATE tablename will stay in a lock state until the
pg_dump is complete?
About 60% of our tables are refreshed daily from our Filemaker database
and we don't care if one or more of these tables are e
* Dearman, Rick <[EMAIL PROTECTED]> [1245 05:45]:
> I have a requirement from my security manager but I can't seem to find a good
> solution.
> So I wondered if someone had done something similar.
>
> We have a web portal and a DB in PostgreSQL (obviously) which contains user
> data.
> The por
You are right again.
I don't want to make permanet changes for this one-time problem.
Yes I find a way to
Use pg_restore to make a text dump script, and then edit that, and then
load it into your newer server.
- pg_restore can output to file not only database. I will edit this file.
Thanks.
Kaloyan
Hi,
I had a crash with my hdd and now I have recovered the database with the
directory
found in lost and found.
Now I get an error, if I try to open the database admin tool PGAdmin III:
cannot open relation pg_amop
What is missing ?
Coul'd this corrected ?
Thanks
Lothar
-
Ah, the tired old 'turn it into a closed product because its BSD' idea.
If
it wasn't so prevalent and pathetic it would be funny.
Other than the fact that ITS HAPPENED ONCE ALREADY (see Illustra)
many moons ago and hasn't managed to derail further development
(quite the opposite, actually - PG gr
On Mon, Dec 20, 2004 at 12:13:31 +,
"Vladimir S. Petukhov" <[EMAIL PROTECTED]> wrote:
> Hi
> Sorry for my English..
>
> I need to organize database structure for saving statistic data for objects.
> I
> have about 24 * 31 * 4 fields (4 month, 31 days, 24 hours) of data for one
> object.
On Mon, Dec 20, 2004 at 17:21:02 -,
"Dearman, Rick" <[EMAIL PROTECTED]> wrote:
> I have a requirement from my security manager but I can't seem to find a good
> solution. So I wondered if someone had done something similar.
>
> We have a web portal and a DB in PostgreSQL (obviously) which c
A much better way to do this, assuming that the database login uses the
actual username/password of the manager in question (if not, see if
this can be arranged securely):
Assume you have a table with the restricted information. Call it
real_data:
CREATE TABLE real_data
(
accountID INTEGE
On Mon, 20 Dec 2004, Marcelo Cid wrote:
suspected the original link you posted:
Mike Mascari
Excuse me. But I dont tried to spoof the vote.
I only copied the link after I voted, so the site returned this link.
They used a 'GET method' for the vote, so the link you sent to the list
had the same vot
Patrick Hatcher <[EMAIL PROTECTED]> writes:
> Curious: Why would a DELETE FROM tablename work while a pg_dump is
> occurring but a TRUNCATE tablename will stay in a lock state until the
> pg_dump is complete?
TRUNCATE requires an exclusive lock on the table.
This is pretty much a no-free-lunch
Ok, this is a real example:
CREATE TABLE account (
val1 BIGINT NULL,
val2BIGINT NULL,
...
dayposSMALLINTNULL, -- Day position
hourpos SMALLINTNULL, -- Hour position
idINT NULL -- Link to the object
);
On Tuesday 21 December 2004 1
Instead of having separate fields for day, hour, ... - why not use
timestamp values?
On Dec 21, 2004, at 3:47 PM, Vladimir S. Petukhov wrote:
Ok, this is a real example:
CREATE TABLE account (
val1 BIGINT NULL,
val2BIGINT NULL,
...
dayposSMALLINTNULL, -- Day pos
I'm trying to install PG 7.4.6 (from Fedora rpms) and
keep getting stuck with psql refusing to do anything
useful. As a simple example:
---
bash-3.00$ createdb postgres
CREATE DATABASE
bash-3.00$ psql
Welcome to psql 7.4.6, the PostgreSQL interactive terminal.
Type:
Steve Wampler wrote:
I don't see any errors when starting with
"service postgresql start" - and if I empty out
/var/lib/pgsql, the command initializes the database
with no errors.
To add a bit more information, and correct the above.
I wasn't seeing any errors because PGLOG was set
to /dev/null in
Steve Wampler <[EMAIL PROTECTED]> writes:
> I'm trying to install PG 7.4.6 (from Fedora rpms) and
> keep getting stuck with psql refusing to do anything
> useful.
Are you running with SELinux enforcement mode on? We've just isolated
some problems there --- see
https://bugzilla.redhat.com/bugzilla
"Jim C. Nasby" <[EMAIL PROTECTED]> writes:
> On Fri, Dec 17, 2004 at 12:07:21PM -0500, Greg Stark wrote:
> >
> > Csaba Nagy <[EMAIL PROTECTED]> writes:
> >
> > > Now, cron isn't exactly part of the OS, is it ?
> >
> > Yes, it is.
>
> Depends on how you define 'OS'. Linux is nothing but a kern
Tom Lane wrote:
Steve Wampler <[EMAIL PROTECTED]> writes:
I'm trying to install PG 7.4.6 (from Fedora rpms) and
keep getting stuck with psql refusing to do anything
useful.
Are you running with SELinux enforcement mode on? We've just isolated
some problems there --- see
https://bugzilla.redhat.co
[EMAIL PROTECTED] (Patrick Hatcher) writes:
> Curious: Why would a DELETE FROM tablename work while a pg_dump is
> occurring but a TRUNCATE tablename will stay in a lock state until the
> pg_dump is complete?
DELETE FROM merely has to mark all the tuples as dead, which requires
no control over th
On Tue, Dec 21, 2004 at 20:47:31 +,
"Vladimir S. Petukhov" <[EMAIL PROTECTED]> wrote:
> Ok, this is a real example:
>
> CREATE TABLE account (
> val1 BIGINT NULL,
> val2BIGINT NULL,
> ...
>
> dayposSMALLINTNULL, -- Day position
> hourpos SMALLINTNU
On Tue, Dec 21, 2004 at 03:42:22PM -0500, Greg Stark wrote:
>
> "Jim C. Nasby" <[EMAIL PROTECTED]> writes:
>
> > On Fri, Dec 17, 2004 at 12:07:21PM -0500, Greg Stark wrote:
> > >
> > > Csaba Nagy <[EMAIL PROTECTED]> writes:
> > >
> > > > Now, cron isn't exactly part of the OS, is it ?
> > >
>
On Tuesday 21 December 2004 21:21, Bruno Wolff III wrote:
> On Tue, Dec 21, 2004 at 20:47:31 +,
>
> "Vladimir S. Petukhov" <[EMAIL PROTECTED]> wrote:
> > Ok, this is a real example:
> >
> > CREATE TABLE account (
> > val1 BIGINT NULL,
> > val2BIGINT NULL,
> > ...
> >
>
Sorry to duplicate this but I thought some of you might not have noticed
my other email about this. I have registered pgJob at pgfoundary
(http://pgfoundry.org/projects/pgjob/). The intention of pgJob is to
provide a means to run SQL commands in a database at specified
intervals. There is a mailing
I've been contracted to provide 3 days of PostgreSQL training, and I'm
wondering if anyone has curriculum they'd like to share with me, or
suggestions for course materials. This course is targeted at database
experts who want to come up to speed on PostgreSQL, so it will deal
mostly with installati
On Wed, Dec 22, 2004 at 00:16:06 +,
"Vladimir S. Petukhov" <[EMAIL PROTECTED]> wrote:
> On Tuesday 21 December 2004 21:21, Bruno Wolff III wrote:
> > On Tue, Dec 21, 2004 at 20:47:31 +,
> >
> > "Vladimir S. Petukhov" <[EMAIL PROTECTED]> wrote:
> > > Ok, this is a real example:
> > >
> >
"Jim C. Nasby" <[EMAIL PROTECTED]> writes:
> I am being paid for this training, so keep that in mind if you're a real
> GPL zealot. ;)
Why would a GPL zealot care if you are being paid to provide training?
---(end of broadcast)---
TIP 5: Have you c
On Tuesday 21 December 2004 22:00, Bruno Wolff III wrote:
> On Wed, Dec 22, 2004 at 00:16:06 +,
>
> "Vladimir S. Petukhov" <[EMAIL PROTECTED]> wrote:
> > On Tuesday 21 December 2004 21:21, Bruno Wolff III wrote:
> > > On Tue, Dec 21, 2004 at 20:47:31 +,
> > >
> > > "Vladimir S. Petukhov
On Tue, Dec 21, 2004 at 04:13:46PM -0600, Doug Quale wrote:
> "Jim C. Nasby" <[EMAIL PROTECTED]> writes:
>
> > I am being paid for this training, so keep that in mind if you're a real
> > GPL zealot. ;)
>
> Why would a GPL zealot care if you are being paid to provide training?
Maybe 'GPL zealot
Or anything else for that matter?
The GPL specifically allows you to charge even for the software (for
other people's work, in fact); the catch is that you can't prevent the
people you sell it to from giving it away for free, so what's the point
;-) (although companies like redhat and mandrake
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Why would it be?
The only real advantage to what you are suggesting would be slightly
reduced disk space usage, and just maybe *very* slightly improved
performance on low-memory servers, or servers with very slow disks, but
I find it unlikely that th
(Originally posted on -hackers, but they were too busy or not interested.)
I am thinking about attempting to code this if it is not too difficult
for a newbie, but I wanted to see if anyone has any input or ideas first.
Goal: on a prduction server, to gradually shrink a table (no matter how
large
I absolutly agree with you, thank.
On Tuesday 21 December 2004 22:50, you wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> Why would it be?
>
> The only real advantage to what you are suggesting would be slightly
> reduced disk space usage, and just maybe *very* slightly improved
> per
BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
The release candidate for DBD::Pg 1.40 is now available. This release
has a very large number of changes and new features, so if you use
Perl and PostgreSQL, please try it out on your platform. A tarball can
be found here:
http://www.gtsm.com/dbd
Hi all,
I'm using psql 8.0.0 on a client's site who's running win server 2003.
We've had him on beta 3 for some time, and no problems at all (yes, in a
sense, he is a beta tester as well, but doesn't know it!). Today I tried to
upgrade the db to RC1 and had some problems.
Remote clients connect
A. Mous wrote:
Hi all,
I'm using psql 8.0.0 on a client's site who's running win server 2003.
We've had him on beta 3 for some time, and no problems at all (yes, in a
sense, he is a beta tester as well, but doesn't know it!). Today I tried to
upgrade the db to RC1 and had some problems.
Remote cli
On Wed, Dec 22, 2004 at 01:24:57 +,
"Vladimir S. Petukhov" <[EMAIL PROTECTED]> wrote:
> Yes, of course, this is example only.
> But relation between tables is not important now...
It is important for design. You should use a normallized design initially
and consider denormalized designs if y
On Tue, Dec 21, 2004 at 17:50:06 -0500,
"Frank D. Engel, Jr." <[EMAIL PROTECTED]> wrote:
>
> Also, given the amount of data you are talking about, and assuming that
> you are inserting all of this data in one big lump, you may wish to
> VACUUM FULL after doing your INSERTs (not after each one,
Hey there Tom thanks for the answer.
However, as you saw I wrote this early in the morning and forgot an
important piece of information:
The table at the time of the truncate was not being dumped. I could see
in pg_stat_activity that it was chugging away at one of the 63M row
tables I have.
Do
On Tue, Dec 21, 2004 at 07:09:39PM -0500, Paul Tillotson wrote:
> To use this system one would do this:
>
> (1) VACUUM KEEP_EARLY_FREE_PAGES mybloatedtable; -- item (a)
>
> (2) UPDATE mybloatedtable SET foo = foo WHERE ctid > '(n, 0)';
> --move tuples in end of the table to the front.
>
> (3
Can anyone suggest good open source replicator system for pgsql
Thanks
JD
---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
joining column's datatypes do not match
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
check out slony. works great
On Tuesday 21 December 2004 08:56 pm, Jamie Deppeler wrote:
> Can anyone suggest good open source replicator system for pgsql
>
> Thanks
> JD
>
> ---(end of broadcast)---
>
Its been 3 weeks since our first Release Candidate, and we're down to the
finally stretch for Full Release, which is looking good for happening
before the New Year.
A current list of *known* supported platforms can be found at:
http://developer.postgresql.org/supported-platforms.html
We'
Hello,
First some specifics:
OS Linux FC1
1 Gig of Ram
IDE RAID 1 array
AMD 2200 XP
Running both 8.0rc2 and 7.4.6 for testing. The function is a plphp
function. The code for the function is below:
CREATE OR REPLACE FUNCTION get_users(bigint[]) RETURNS SETOF
resolved_users_type AS '
$userData =
On Tue, Dec 21, 2004 at 07:06:00PM -0800, Joshua D. Drake wrote:
> The clients connecting remotely... do they support SSL? It looks
> like to me that you have ssl enabled. I would check the pg_hba.conf
> and the postgresql.conf for your ssl configuration.
Was there a change affecting client certi
The SSL line in postgresql.conf has not been touched. By default it should
be set to false. Also, the pg_hba file is set to allow all clients in, no
authentication (TRUST).
I have not seen this problem in any other installation. Note as well, the
only way to access the server in question is via
46 matches
Mail list logo
