Обсуждение: cannot connect to database

Поиск
Список
Период
Сортировка

cannot connect to database

От
Laurent GALAIS
Дата:
Hello,

I have a problem when trying to connect to a specific database with
postgres.

I'm logged as a user helpdesk group helpdesk

in my environment :

helpdesk@phoenix:~ > env |grep -i pg

PGLIB=/var/lib/pgsql/lib
PGDATA=/helpdesk/dbserver/db/data

PATH=/support/bin:/usr/local/bin:/usr/bin:/usr/X11R6/bin:/bin:/usr/lib/java/bin:/usr/games/bin:/usr/games:/opt/gnome/bin:

/opt/kde/bin:/usr/openwin/bin:.:/var/lib/pgsql/bin:/var/lib/pgsql/bin:/var/lib/pgsql/bin

Postgresql is installed in /var/lib/pgsql - version 6.4.2

In data :

helpdesk@phoenix:/helpdesk/dbserver/db/data > ll
total 56
-rwxr-xr-x   1 helpdesk helpdesk        4 Jan 23 16:18 PG_VERSION
drwxr-xr-x   5 helpdesk helpdesk     4096 Dec  5  1999 base
-rwxr-xr-x   1 helpdesk helpdesk     8192 Dec  5  1999 pg_database
-rwxr-xr-x   1 helpdesk helpdesk     3407 Dec  5  1999 pg_geqo.sample
-rwxr-xr-x   1 helpdesk helpdesk        0 Dec  5  1999 pg_group
-rwxr-xr-x   1 helpdesk helpdesk     5192 Dec  5  1999 pg_hba.conf
-rwxr-xr-x   1 helpdesk helpdesk     8192 Dec  5  1999 pg_log
-rwxr-xr-x   1 helpdesk helpdesk       52 Dec  5  1999 pg_pwd
-rwxr-xr-x   1 helpdesk helpdesk     8192 Dec  5  1999 pg_shadow
-rwxr-xr-x   1 helpdesk helpdesk     8192 Dec  5  1999 pg_variable

In data/base :

drwxr-xr-x   2 helpdesk helpdesk     4096 Dec  5  1999 postgres
drwxr-xr-x   2 helpdesk helpdesk     4096 Dec  5  1999 problem_track
drwxr-xr-x   2 helpdesk helpdesk     4096 Dec  5  1999 template1

A postmaster process is running :

6709 pts/13   S      0:00 postmaster -B 256 -i -D /helpdesk/dbserver/db/data

In this environment I try :

psql template1 Ok

psql problem_track :
Connection to database 'problem_track' failed.
FATAL 1:  Database problem_track does not exist in pg_database

Can someone help me and point me to my mistake

TIA

Laurent



Re: cannot connect to database

От
Einar Karttunen
Дата:
On Mon, 29 Jan 2001, Laurent GALAIS wrote:
>
> psql problem_track :
> Connection to database 'problem_track' failed.
> FATAL 1:  Database problem_track does not exist in pg_database
>
> Can someone help me and point me to my mistake
>
Does the database exist? Have you created it with createdb?
createdb problem_track
psql problem_track

If you have created the database check whether the files from the
database exist.

- Einar Karttunen


Re: cannot connect to database

От
Laurent GALAIS
Дата:

Einar Karttunen wrote:

> On Mon, 29 Jan 2001, Laurent GALAIS wrote:
>
>> psql problem_track :
>> Connection to database 'problem_track' failed.
>> FATAL 1:  Database problem_track does not exist in pg_database
>>
>> Can someone help me and point me to my mistake
>>
> Does the database exist? Have you created it with createdb?
> createdb problem_track
> psql problem_track
>
> If you have created the database check whether the files from the
> database exist.
>
> - Einar Karttunen

Thanks for your reply,

In fact the database problem_track comes with a software.

I have done a createdb, because I though the createdb was already done
because in data/base there is a directory problem_track with a lot of
files - one of these is staff for example :

file staff :  staff: data

I have tried a createdb problem_track : now psql problem_track exists
for pl_sql but no tables are inside whereas they are in
data/base/problem_track.

Any ideas ?

Thank you

Laurent

>
>
>
>


select using index in group by slower as normal fields?

От
Einar Karttunen
Дата:
I have a table called student from which I want to return all students
to an application. If I do a query like SELECT * FROM STUDENT; explain
tells me that the cost is between 0 and 13.44. The result is same if the
table is ordered by fname, lname or class. If I order by id, then
postgresql uses index scan which takes from 0 to 43.09. Are the values
given by explain just bad or why does it appear that sorting a table
is in this case, better with any other than the index key? I have used
VACUUM and VACUUM ANALYZE periodically.

- Einar Karttunen


ekarttun=# \d student
                             Table "student"
 Attribute |   Type   |                     Modifier
-----------+----------+--------------------------------------------------
 id        | integer  | not null default nextval('student_id_seq'::text)
 fname     | char(15) |
 lname     | char(15) |
 class     | char(3)  |
Index: student_pkey
Constraint: (class ~ '^[0-9][0-9][A-Z]$'::text)

ekarttun=# explain select * from student;
NOTICE:  QUERY PLAN:

Seq Scan on student  (cost=0.00..13.44 rows=644 width=40)

EXPLAIN
ekarttun=# explain select * from student order by fname;
NOTICE:  QUERY PLAN:

Sort  (cost=43.49..43.49 rows=644 width=40)
  ->  Seq Scan on student  (cost=0.00..13.44 rows=644 width=40)

EXPLAIN
ekarttun=# explain select * from student order by lname;
NOTICE:  QUERY PLAN:

Sort  (cost=43.49..43.49 rows=644 width=40)
  ->  Seq Scan on student  (cost=0.00..13.44 rows=644 width=40)

EXPLAIN
ekarttun=# explain select * from student order by class;
NOTICE:  QUERY PLAN:

Sort  (cost=43.49..43.49 rows=644 width=40)
  ->  Seq Scan on student  (cost=0.00..13.44 rows=644 width=40)

EXPLAIN
ekarttun=# explain select * from student order by id;
NOTICE:  QUERY PLAN:

Index Scan using student_pkey on student  (cost=0.00..43.09 rows=644
width=40)

EXPLAIN





Re: select using index in group by slower as normal fields?

От
Einar Karttunen
Дата:
On Wed, 31 Jan 2001, Einar Karttunen wrote:
>...

ignore the previous message, I dismissed the sort times...

- Einar