"Mihai Gheorghiu" wrote:
>With createuser I managed to create users like 12345 and John Doe (with a
>space).
>It did not work with CREATE USER.
>Aren't they supposed to do the same?
Such users need to be double-quoted -- which is how createuser does it:
QUERY="CREATE USER \"$NewUser\""
.
I need to store very large graphs structures in
postgres. The graphs are close to 20GB when in
flatfile format. I first tried using an adjacency
list representation, i.e.,
graph (source INT8 PRIMARY KEY, dest INT8[]);
but operating on the array type seems a bit
inflexible. I took a look at t
On Fri, Aug 24, 2001 at 09:53:00PM +, uncleleo wrote:
> I am attemping to create multiple databases with Postgresql ver. 7.0.3
> running on Mandrake 8.0 rpm. The tool that I am using is Pgadmin ver 7.1.0.
>
> Can someone tell me how I can connect to different databases in a single
> select s
=?iso-8859-1?Q?Daniel_=C5kerud?= <[EMAIL PROTECTED]> writes:
> I want to create a sequence with a higher value than an int4.
> Please explain how to do this, if possible.
There's no way in current releases.
> If not, this is certainly an issue that has to be targeted real soon, or?
Sequences ar
in looking for a way to abstract the generation of formatting
methods based on field type (dates=center & formatted; numerics
right-flush; text left-flush, etc) for web pages i've run into a
gap i don't know how to close:
using perl (5.005_03) and DBI (1.13, which refers to
DBD/Pg
On Wed, 5 Sep 2001, Gabriel Fernandez wrote:
> Hi,
>
> What does it mean the MATCH FULL parameter in a foreign key specfication
> ?
>
> It seems it is only valid for a multi-column foreign key: What's a
> multi-column foreign key ?
It's only different on multi-column fks, it's valid but unint
On Wed, Aug 29, 2001 at 10:44:56AM +0200, Stan wrote:
> How can I change the sequence of characters in PostgreSQL character indexes
create index some_ix
on some_tbl ( substr(some_fld,3,1) || substr(some_other_fld,12,5) );
but if you look hard enough, you'll usually find a better model
by
Jerry Asher <[EMAIL PROTECTED]> writes:
> Can anything be done to recover the pg_control file?
I've already been talking to Jerry off-list, and I trust he's managed
to recover his data --- but for the archives, the answer to this is to
hack up contrib/pg_resetxlog to force whatever data you need
I use PostgreSQL 7.1.3 on both Linux 2.4 and Solaris 8 with my multi
threaded application. The application has 16 threads and each thread
opens a connection to PostgreSQL. Under Linux, this works just
fine. Under Solaris, 11 threads succeeds in connecting to the
database, but the rest fails with:
On Sun, 02 Sep 2001 20:01:27 -0400, Randall Perry wrote:
> I've got an if statement that checks if a boolean value is true:
>
> if ($cust_data->{'hold'} eq 't')
>
> But perl will not evaluate the value. $cust_data->{'hold'} is taken from a
> PgSQL boolean field (either t or f). If I use the co
Hi,
What does it mean the MATCH FULL parameter in a foreign key specfication
?
It seems it is only valid for a multi-column foreign key: What's a
multi-column foreign key ?
I've tried to search this in Bruce's book but I have a lot of problems
to connect to it at http://www.postgresql.org/doc
I want to create a sequence with a higher value
than an int4.
Please explain how to do this, if
possible.
If not, this is certainly an issue that has to be
targeted real soon, or?
---Daniel Åkerud
[ Don't underestimate the power of stupid people in
large groups]
hi, i'm having trouble with the BigDecimal bug in the jdbc driver using
jbuilder 4 and wonder if some body has a compiled version of the jdbc
with the patch applied and if i can download it
thanks in advance and best regards.
---(end of broadcast)
playpen=# create table jm(
playpen(# jid int NOT NULL,
playpen(# mid int ,
playpen(# UNIQUE(jid, mid)
playpen(# );
NOTICE: CREATE TABLE/UNIQUE will create implicit index 'jm_jid_key' for table 'jm'
CREATE
Why isn't the index created called 'jm_jid_mid_key' ?
--
Joseph Shraibman
[EMAIL PROTEC
Oh, Ok. Now I understand. It seems that there is no solution
using SQL without using inheritance. I had some Ideas but postgres
has weird sematics (not conforming SQL99) using inheritance and so
it didn't work. But you should not - as you suggested - store the
tablenames in a "assocciation" tab
Gabriel Fernandez wrote:
> Hi,
>
> What's the difference between NOT DEFERRABLE and INITIALLY IMMEDIATE for
> a FOREIGN KEY specification ?
>
> It seems they are both defining the same: the moment the constraint will
> be checked: the instant the statement is processed or the end of
> transaction.
(Sorry for reposting...)
> You said you enter some rows, but how many rows got returned by your
query?
You're right, it does make all the difference. With only a few rows,
indexing is somehow used in the first case only (for my example), but with
larger tables index scan is used in both ways.
Somehow in upgrading from 7.1.2 to 7.1.3 (and with much operator error
along the way), my database became corrupted.
I believe it is only the pg_control file that was damaged.
The database was completely shutdown at the time.
Is there anyway to create a new pg_control file, based on the conten
* [EMAIL PROTECTED] (Randal L. Schwartz) wrote:
|
| If you'd just stop saying things that can't be backed up, I wouldn't
| have to keep responding.
;-)
| Where is your proof that mod_perl is "slower than most any j2ee
| application"?
I don't like to generalize either. I've been on two projects
> "Gunnar" == Gunnar Rønning <[EMAIL PROTECTED]> writes:
Gunnar> In the end there is however no proof to claim that Java applications
Gunnar> are faster than mod_perl applications.
I'll settle for that. Most of the time I've seen benchmarks, it's
been more the skillset of the programmers at
With createuser I managed to create users like 12345 and John Doe (with a
space).
It did not work with CREATE USER.
Aren't they supposed to do the same?
Thank you all,
Mihai Gheorghiu
---(end of broadcast)---
TIP 1: subscribe and unsubscribe comm
Thanks to all who suggested indexing; I added an index to
the running database, and delete performance immediately
increased by a factor or 10.
The problem now is that when I "vacuum verbose analyze", the
pg_largeobject system table takes *forever* (tens of minutes).
A new postmaster seems to be
What does your data look like? If you have a lot of duplicate id's, a
sequential scan may be better than an index scan.
If you are not sure if this is the case, try:
SELECT id, count(*) AS count FROM test GROUP BY id ORDER BY count DESC
LIMIT 50;
This should show you the top 50 most duplicate
Hi everyone,
I would like to take advantage of postgresql's ability to store arrays
within records, but I wanted to make sure that I store them in the most
efficient way so my performance is optimized. I want to store
one-dimensional arrays of float values, each with about 6000 elements. I
ne
24 matches
Mail list logo
