Hello Tom,
Late answer, but answer :-) :
Finally, it was a very strange hardware problem, where a very small part
of RAM was defect but kernel never crashed.
I had also a very strange behavior when verifying rpm packages with rpm
-V. First I had the harddisk under suspicion. But then I flushe
Hello,
Finally I used a function below which works well. Only one problem is
left: It polutes the buffer cache because of the cursor. Any idea to get
rid of this behavior?
BTW: WINDOWING FUNCTION of 8.4 should help but noone could provide an
examples how this could work. Any further comments
"Brendan Hill" writes:
> Makes sense to me. Seems to be happening rarely now.
> I'm not all that familiar with the open source process, is this likely to be
> included in the next release version?
Can you confirm that that change actually fixes the problem you're
seeing? I'm happy to apply it i
Sam Mason writes:
> Hum, why is PG doing an (unchecked) atoi on the user specified port
> rather than leaving it up to getaddrinfo to resolve the port? It would
> seem to require changing UNIXSOCK_PATH to accept a string as the "port
> number", which is probably a bit much of a change.
> The inc
On Sat, Sep 26, 2009 at 8:26 PM, Martin Gainty wrote:
> i prefer cygwin as all the unix commands are available but mkuser and
> groupadd arent up yet..ugh
You might be happier with VMs running BSD or some linux flavor. If
you have enough memory and 64 bit windows it should run well enough.
Being
Martin Gainty wrote:
whats *supposed to* happen is
RUNAS will run any command (the question is are we working with
Windows Native Binaries or cygwin?)
with regards to cygwin
i can build postmaster.exe postgres.exe initdb.exe, libpgport_srv.dll
and all the binaries using gcc and ld
cygwin co
i prefer cygwin as all the unix commands are available but mkuser and groupadd
arent up yet..ugh
i have a bunch of apps already running..webapp
servers..tomcat..glassfish..weblogic
and 2 other databases running and because I need god access (I cant run any of
these as postgres user)
here is t
On Sat, Sep 26, 2009 at 8:10 PM, Martin Gainty wrote:
> partial victory but
> im baffled i started all commands with RUNAS /user:postgres
> cd /postgres/pgctl/bin
> RUNAS /user:postgres "initdb.exe -D /data"
> RUNAS /user:postgres ""pg_ctl.exe" -D "/data" -l logfile start"
> RUNAS /user:postgres "
So, did you delete the postgres user?
Go here:
http://www.postgresql.org/docs/8.4/static/app-postgres.html
and look for single user mode. It's likely what you need.
Not surprised you can't get in with psql, you seem to not have a super
user account (or possibly any account)
2009/9/26 Martin Gai
cant get in at all with psql
netstat shows the listener is up on 5432
trying with pgAdmin3
Martin Gainty
__
Verzicht und Vertraulichkeitanmerkung/Note de déni et de confidentialité
Diese Nachricht ist vertraulich. Sollten Sie nicht der vorgesehene E
So, did you delete the postgres user? What happens if you bring up a
backend in single user mode? Can you see the users then with \du?
2009/9/26 Martin Gainty :
> postgres
> ?
> Martin Gainty
> __
> Verzicht und Vertraulichkeitanmerkung/Note de déni et
2009/9/26 Martin Gainty :
> All-
>
> when starting postgres what does this error mean exactly?
> psql: FATAL: role "postgres" does not exist
>
> what is the exact solution for psql: FATAL: role "postgres" does not exist
> error?
That's not an error starting postgresql, it's an error starting psq
All-
when starting postgres what does this error mean exactly?
psql: FATAL: role "postgres" does not exist
what is the exact solution for psql: FATAL: role "postgres" does not exist
error?
thank you
Martin Gainty
__
Verzicht und Vertraulichkeitan
> Just to make things clear, is this the same question and situation that
> Jerzy Bialas brought up in another post?
>
> http://archives.postgresql.org/pgsql-general/2009-09/msg00922.php
Yes, I didn't know he posted it earlier.
Sorry for the confusion and thanks for response.
Regards,
Krzysztof
Thank you for your response; I understand the information provided was
somewhat limited; I am happy to provide a bit more though: I notice you guys
have quite experience modeling data...
What I am trying to do is: building views on a base table, extended by one
or more columns, extracted (hence the
On Sat, Sep 26, 2009 at 1:23 PM, Gerhard Wiesinger wrote:
> On Sat, 26 Sep 2009, Tom Lane wrote:
>
>> Gerhard Wiesinger writes:
>>>
>>> Research on the pg_bufferache showed that a lot of pages don't have any
>>> assigned relfilenodesin pg_class, even when they are dirty (in this case
>>> inserts
Thank you for your response; I understand the information provided was
somewhat limited; I am happy to provide a bit more though: I notice you guys
have quite experience modeling data...
What I am trying to do is: building views on a base table, extended by one
or more columns, extracted (hence the
What I ended up doing to work around not seeing anything besides your
active database for my own tests was write a simple shell script that
connected to all of the database in succession, saving everything in the
catalog that pg_buffercache needs to join against to a table. Then I
joined again
On Saturday 26 September 2009 12:49:03 pm Krzysztof Barlik wrote:
> Hi Tom,
>
> > If that's not what you are talking about, you need to be more
> > specific about what you are talking about.
>
> you are right, I was not specific enough.
>
> I meant a situation where returned rowset
> may have diffe
On Saturday 26 September 2009 1:10:20 pm Pavel Stehule wrote:
> Hello
>
> 2009/9/26 Adrian Klaver :
> > On Friday 25 September 2009 9:52:09 am Pavel Stehule wrote:
> >> Hello
> >>
> >> this isn't possible now. All what you want are limited by SELECT
> >> statement in PostgreSQL. Simply PL knows onl
On Thu, Sep 24, 2009 at 06:28:28PM +0200, InterRob wrote:
> Dear List,
> I am trying to implement the following:
>
> In a database I wish to implement a GENERIC datamodel, thus on a
> meta-level.
That's not a very bright idea, even though it seems so when you first
think of it.
Relational databa
2009/9/26 Krzysztof Barlik :
> Hi Tom,
>
>> If that's not what you are talking about, you need to be more
>> specific about what you are talking about.
>
> you are right, I was not specific enough.
>
> I meant a situation where returned rowset
> may have different set of column names depending
> on
Hello
2009/9/26 Adrian Klaver :
> On Friday 25 September 2009 9:52:09 am Pavel Stehule wrote:
>> Hello
>>
>> this isn't possible now. All what you want are limited by SELECT
>> statement in PostgreSQL. Simply PL knows only SELECT statement and
>> structure of result have to be strongly static - b
On 26 Sep 2009, at 19:57, Tom Lane wrote:
I think you want INSERT ... RETURNING ... INTO some-plpgsql-variable;
regards, tom lane
On 26 Sep 2009, at 19:56, Adrian Klaver wrote:
See:
http://www.postgresql.org/docs/8.4/interactive/plpgsql-
statements.html#PLPGSQL-S
On Friday 25 September 2009 9:52:09 am Pavel Stehule wrote:
> Hello
>
> this isn't possible now. All what you want are limited by SELECT
> statement in PostgreSQL. Simply PL knows only SELECT statement and
> structure of result have to be strongly static - because it is based
> on static execution
Hi Tom,
> If that's not what you are talking about, you need to be more
> specific about what you are talking about.
you are right, I was not specific enough.
I meant a situation where returned rowset
may have different set of column names depending
on data in the database, not on explicit user
Hi,
I have a question - is it possible to rename
columns in rowset returned from plpgsql function
delared as 'returns table(...)' ?
It seems it's not possible, even using
'as "column_name"' in internal query and
column names are always the same as in
'table(...)' definition.
If not, could it be
On Sat, 26 Sep 2009, Tom Lane wrote:
Gerhard Wiesinger writes:
Research on the pg_bufferache showed that a lot of pages don't have any
assigned relfilenodesin pg_class, even when they are dirty (in this case
inserts are done).
There should never be any buffered pages for nonexistent tables.
Dear my friends...
I've installed postgresql-server on OpenSuSE11.1 successfully but I
can't connect to it from psql.
I did these steps:
"
1. I created a new opensuse11.1-linux login account + its password
(username: ivia) with YaST2;
2. i...@sussy:~> su postgres -c psql postgres
Passwort:
Dies
> Can anyone help me with this?
>
> In what encoding the function PQgetvalue() return?
> ASCII? UTF8?
It returns in whatever client encoding you are using:
test=> SHOW client_encoding;
client_encoding
-
SQL_ASCII
(1 row)
--
Bruce Momj
Iain Barnett writes:
> I've written a straightforward insert function, but using the
> RETURNING keyword for the first time. If I try running the test case
> I get the error:
> ERROR: query has no destination for result data
> CONTEXT: PL/pgSQL function "nonauth_users_insert_new_udf" line 7
On Saturday 26 September 2009 11:04:42 am Iain Barnett wrote:
> I've written a straightforward insert function, but using the
> RETURNING keyword for the first time. If I try running the test case
> I get the error:
>
> ERROR: query has no destination for result data
> CONTEXT: PL/pgSQL function
Gerhard Wiesinger writes:
> Research on the pg_bufferache showed that a lot of pages don't have any
> assigned relfilenodesin pg_class, even when they are dirty (in this case
> inserts are done).
There should never be any buffered pages for nonexistent tables. The
method you are using will onl
On Sep 26, 2009, at 11:29 AM, Christophe Pettus wrote:
Of course, one can create a user that has no useful privileges on a
particular database, but is there a way of limiting a user to being
able to log into only a single one of, or a subset of, the databases
in a cluster?
Like, say GRAN
Of course, one can create a user that has no useful privileges on a
particular database, but is there a way of limiting a user to being
able to log into only a single one of, or a subset of, the databases
in a cluster?
--
-- Christophe Pettus
x...@thebuild.com
--
Sent via pgsql-general
John R Pierce píše v so 26. 09. 2009 v 10:57 -0700:
> Zdenek Kotala wrote:
> >> I asked because we're working on an inhouse app using postgres, the team
> >> wants to use 8.4... I had previously packaged 8.3.7 in solaris package
> >> format only to have it come out as a patch a month or two la
I've written a straightforward insert function, but using the
RETURNING keyword for the first time. If I try running the test case
I get the error:
ERROR: query has no destination for result data
CONTEXT: PL/pgSQL function "nonauth_users_insert_new_udf" line 7 at
SQL statement
I'm not s
Zdenek Kotala wrote:
I asked because we're working on an inhouse app using postgres, the team
wants to use 8.4... I had previously packaged 8.3.7 in solaris package
format only to have it come out as a patch a month or two later, so I
was hoping to save myself the redundant effort.
You
Hello,
Research on the pg_bufferache showed that a lot of pages don't have any
assigned relfilenodesin pg_class, even when they are dirty (in this case
inserts are done).
SELECT
relname IS NULL AS relame_is_null,
isdirty,
COUNT(*) AS count
FROM
pg_buffercache b
LEFT OUTER JOIN pg_clas
John R Pierce píše v so 26. 09. 2009 v 09:44 -0700:
> >> I'm curious if anyone has a clue how long before Sun releases PG 8.4.x
> >> in Solaris 'patch' format ...
> >>
> >
> > Currently S10 is shipped with three version (8.1, 8.2, 8.3) of
> > PostgreSQL. There is no plan to integrate PG8.4
John R Pierce píše v út 15. 09. 2009 v 19:53 -0700:
> anyone on the PG team privvy to Sun Solaris patches for postgres? like
> patch 138826-04 is Postgres 8.3.7 for Solaris10 Sparc... 138827-04 is
> the same for Sol10 x86...
What is wrong on this? It is standard Solaris way how to deliver
VANOLE, MICHAEL J (ATTSI) píše v pá 11. 09. 2009 v 11:58 -0500:
> Hi,
>
> We are running Pg 8.2.6 on sparc-sun-solaris2.9 compiled by gcc 3.4.6.
>
> We are planning a server software upgrade to Solaris 10 (binaries and
> kernel, etc).
>
> Has anyone done this, and were there problems when tryi
Krzysztof Barlik writes:
> I have a question - is it possible to rename
> columns in rowset returned from plpgsql function
> delared as 'returns table(...)' ?
Sure ...
regression=# create function foo() returns table (a int, b text) as
regression-# $$ values (1,'one'), (2,'two') $$ language sql;
Zdenek Kotala wrote:
John R Pierce píše v út 15. 09. 2009 v 19:53 -0700:
anyone on the PG team privvy to Sun Solaris patches for postgres? like
patch 138826-04 is Postgres 8.3.7 for Solaris10 Sparc... 138827-04 is
the same for Sol10 x86...
What is wrong on this? It is standard S
Dear Sam.
Thank you very much for your help.
You've solved my several problem. But now I'm stucked on another
problem which I know easy to anybody else but difficult as a
postgres starter like me.
I did these steps:
"
1. I created a new opensuse11.1-linux login account + its password
(username:
David Chell wrote on 26.09.2009 06:31:
This is a clean install, I’ve never installed postgresql on the is
machine before, I have successfully installed it on another machine
which is a Windows Server 2008 x64, but whenever I try to install it in
the Windows XP (32) laptop I get the following
Ricky Tompu Breaky wrote:
>
> Dear my friends...
>
> I've installed postgresql-server on OpenSuSE11.1 successfully but I
> can't connect to it from psql.
>
> I did these steps:
> "
> 1. I created a new opensuse11.1-linux login account + its password
> (username: ivia) with YaST2;
> 2. i...@sussy
47 matches
Mail list logo
