y the database and ran initdb again. It says 'postgres' not
in pg_shadow.
My pg_hba.conf has local all trust.
Any clues out there?
--
Peter Eisentraut
PathWay Computing, Inc.
est' it was that
came out of nowhere.
I am not too sure about how I fixed it, but I am pretty sure it was the
user IDs.
--
Peter Eisentraut
PathWay Computing, Inc.
sword "file"?)
--
Peter Eisentraut
PathWay Computing, Inc.
have all access to it anyway.
PostgreSQL 6.5.0 on i686-pc-linux-gnu, compiled by gcc 2.7.2.3
--
Peter Eisentraut
PathWay Computing, Inc.
I just want to verify this: when I update between the said versions, can I
just keep the data directory around (no need to dump/restore)? Would that
eliminate the need to run initdb?
It seems like this could be made slightly clearer in the documentation.
--
Peter Eisentraut
PathWay Computing
the $PGDATA/base directory using
regular file backup methods. That wouldn't be a dump, so it depends on
your particular needs.
--
Peter Eisentraut
PathWay Computing, Inc.
On Tue, 3 Aug 1999, Dmitry Morozovsky wrote:
> On Mon, 2 Aug 1999, Peter Eisentraut wrote:
>
> PE> > How can I organize automated database dumping for passord protected
> PE> > database?
> PE> If you're implying something like "How do I get around en
g_hba.conf(5)), otherwise the system will let you with any password.
In that regard the -u switch is kind of flawed, because prompting for
username and password are really two different things. But that will be
fixed in the next version.
-Peter
--
Peter Eisentraut Sernanders
tions that would drop the rules?
>
>
> Sincerely Yours,
>
> Jacques Dimanche
> VP of Research and Development
> Tridel Technologies, Inc.
>
>
>
>
--
Peter Eisentraut Sernanders vaeg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
>
> Best regards
>
>
> ****
>
>
--
Peter Eisentraut Sernanders vaeg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
tem should take care of; PostgreSQL
won't. Incidentally, I believe that most file systems won't do this, but
something like XFS (SGI) will.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
od database, but its too slow...
There is no such thing as a free lunch ...
--
Peter Eisentraut Sernanders vaeg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
and is it posible at all?!
There's no really good way without making your users superusers. That is
admittedly a deficiency.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
something in
C that does this.
-Peter
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
On 1999-12-11, [EMAIL PROTECTED] mentioned:
> I need a logging utility for postgresql server.
This will (probably) be addressed in WAL (write-ahead logging(?)), due to
appear eventually but, as we found out the other day, not so soon :( .
--
Peter Eisentraut Sernanders väg
On 1999-12-11, lec mentioned:
> How can I find out how many users/who are currently connected to a
> database?
The ps output on the server machine will help.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/p
which you can't), you'll never be able to drop it cleanly either.
I'll submitt a corrected version of this for the next release.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
's not in the RPMs though. (boo)
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
On 1999-12-16, [EMAIL PROTECTED] mentioned:
> On Thu, 16 Dec 1999, Peter Eisentraut wrote:
> > Actually such a script, called vacuumdb, with exactly those options is
> > included in the distribution and should be installed in your bin dir right
> > next to psql.
>
>
o urla di sudore... che schifo...
> --
> Michel Morelli [EMAIL PROTECTED]
> http://www.ziobudda.net
> Italian Linux FAQ http://faq.ziobudda.net/
> Italian Linux Press http://www.ziobudda.net/ILP/
>
&g
rincipia PHONE: (314)-434-2100
> 13201 Clayton RoadFAX: (314)-275-3538
> St. Louis, MO 63131 INTERNET: [EMAIL PROTECTED]
>
>
>
>
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
7;-L$(exec_prefix)' to the link
> line. I don't know the syntax for that, but you get the idea.
In my view that's not reasonable, but anyway. I guess you'd have to do
--with-tcl=/where/it/is
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
"show
> transaction isolation level" tells me that it is still "read commited".
> Any Hints ?
>
> Is there a possibility to have transaction level setted to serializable
> as default level?
>
> Many thanks in advance
>
> Norbert
>
>
>
&g
u are using ident "authentication" then you can do that by setting up
a different mapping file for each database. See pg_hba.conf for
details. If you're using password authentication you have to set up a
different password file for each database. See again pg_hba.conf.
--
Peter Eisent
ome documention on this is greatly appreciated.
>
> thanks
> -mike
>
> mike focosi
> senior applications programmer
> golden dome media
> http://www.gdp.net
> http://www.surfmichiana.com
> http://www.wndu.com
>
>
>
>
>
--
On Tue, 29 Feb 2000, Andrei N.Sobchuck wrote:
> Is it possible to make postgresql to use index scan when it calculate
> MIN(id).
Not currently, but it's on the todo list.
--
Peter Eisentraut Sernanders vaeg 10:115
[EMAIL PROTECTED] 75262 U
lo_import and lo_export
functions.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
N command failed. Explanation from backend: 'ERROR:
> parser: parse error at or near "transaction"
> '.
> pg_dump failed on cqn, exiting
> ========
--
Peter Eisentraut Sernanders väg 10:
transferring them by hand rather than
bothering too much with getting pg_dump[all] to work.
Of course, temporarily disabling the password authentication would work as
well.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
en I now try to access the database using:
>
> [postgres@10 bin]$ psql -u dpstatistics
> username: dpstats
> password: dpstats
Spelling problems?
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED]
Andre Antonio Parmeggiani writes:
> => create table user (user text);
> ERROR: parser: parse error at or near "user"
USER is a reserved word in SQL92. To use it, enclose it in double quotes.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED]
o into the tree before the 7.0 release is cut.
[EMAIL PROTECTED]
If you don't want to receive messages on that list subscribe to
[EMAIL PROTECTED] and you can post to any list without
receiving traffic. Send mail to [EMAIL PROTECTED] to
subscribe. ('subscribe' in body)
--
Peter E
terpan writes:
> (Postgres doesn't work with NT4 no?)
Sure it does. (With cygwin library.)
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
> select * from userdata wher name = 'John'
>
> select * from userdata wher name = 'joHN'
>
> will fetch me the same results.
select * from userdata where lower(name) = 'john'
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PR
directory and pg_pwd file are not world-readable, so that
shouldn't be a problem. I agree that the stuff should be encrypted anyway
but that's a bit of work for a variety of other reasons.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
default, users do not have write access to databases
> they did not create."
The documentation is wrong. You can currently create anything in any
database you have access to. We're currently discussing a way to improve
that in 7.1. Care to share your ideas?
--
Peter Eisentraut
that.)
> can I restrict access with read-write vs. read-only privledges?
Read and write privileges are done on a per table basis via grant and
revoke. See manual entries.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
same registers... what i doing?
select * from books where lower(titulo) like '%perl%';
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
ut, depending on
your application. Alternatively you could try to implement your own data
type based on text but make the = operator case insensitive, but that will
probably create all sorts of extra trouble (indexes come to mind).
> > select * from userdata where lower(name) = 'j
On Tue, 28 Mar 2000, Walter M. Yuan wrote:
> >PQsendQuery() -- query is too long. Maximum length is 16382
> >query buffer max length of 16384 exceeded
>
> Can anyone give me some quick advice on how to change this setting?
Upgrade to 7.0. There isn't a way to change it
On Thu, 6 Apr 2000, Tressens Lionel wrote:
> Is it possible to change the owner of a db after having created it ?
Update the pg_database table in accordance with the user ids listed in
pg_users.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTEC
27;timestamp' column, I got
> this error and creating failed: 'Cannot find a default operator class
> for type 1296.'
Indexes don't work for timestamp in 6.5. Use datetime, or better yet
upgrade to 7.0 which should be out in a few weeks. (jinx)
--
Pe
On Wed, 3 May 2000, brianb wrote:
> Ok, thanks. But why are they there in the first place, then?
A backend crash during sorting
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
ED integer in PostgreSQL ?
> > E.g. how should I describe type of field to have
> > full range of 32 bits? If I describe filed as INT4,
int8/bigint
> > I'll get SIGNED integer, insn't it?
Per SQL, all numbers are signed.
--
Peter Eisentraut Sernander
> > Again, a suggested change?
This is probably what you want:
http://research.cys.de/autoconf-archive/Miscellaneous/ac_func_accept_argtypes.html
I've been meaning to clean up the autoconf mess soon anyway, so I can take
a look at integrating this.
--
Peter Eisentraut
eferred, backing up the PG files or doing some sort of sync?
For backup use pg_dump, for synchronization you're currently out of luck.
> 4. Does anyone have any advice on how often to run VACUUM? Any rules of
> thumb, etc?
If there's a lot of write activity in your dat
trust
> host all0.0.0.0 0.0.0.0 reject
> host all192.168.200.0255.255.255.0 password
Because a mask of 0.0.0.0 matches every host, so the reject kicks in. The
logic here is that
({host entry} XOR {actual host}) AND {mask entry}
must be 0 for a r
> 3.Are there any plans in making security system more
> flexible?
Sure. I'm planning some work on that very soon. Any suggestions you might
want to contribute?
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
Wallingford, Ted writes:
> I am using 6.3 in this case.
I'm sorry but that is pre-historic era around here and no one really
remembers what the problems might have been back then (other than that
they were surely plenty). Upgrading might be your best bet on all fronts.
--
Peter Ei
get in a
repair things. It also looks like a race condition waiting to happen.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
Giving more details on "failure" might help though.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
ess still use the password in clear in pg_shadow . Is there an
> option , i've not set correctly ?
Perhaps pgaccess is not connecting via `local', but through TCP/IP
localhost?
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
you use some authentication
mechanism.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
Hassanuri Bin Shah writes:
> does anyone knows how to change the currency symbol in 'money' type ?
>
> from '$' to 'RM'
`Money' has been deprecated for quite a while now. (The data type, not the
concept :) Use numeric for storage and to_char
HENAFF Mari-mai FTRD/DMI/LAN writes:
> Is it possible to configure PostgreSQL, so that it is not accent mark
> sensitive when making conditions upon strings on the queries ?
No, you'll have to do that yourself. A simple tr 'âàáéÿ' 'aaaey' etc.
could possibly do
on per processor.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
HENAFF Mari-mai FTRD/DMI/LAN writes:
> And is it possible to make a query containing an ORDER BY that is not case
> sensitive ?
Option 1:
SELECT * FROM table ORDER BY upper(field);
Option 2:
Build with --enable-locale. Most locales sort a, A, b, B, etc.
--
Peter Eise
is one processor per process, that is, connection.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
On Tue, 13 Jun 2000, Meibell Contreras wrote:
> Anyone knows if postgres has a bigger data type than money... when you
> use this; (not with dollar definition); the money data type is not
> enough...
`numeric'. And use to_char() for formatting if you like.
--
Pe
Haroldo Stenger writes:
> Besides this, I wonder if db replication could be done by reaplying
> backend logs.
Yup, if we only had backend logs...
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/p
t; How can I remove a user from some group?
ALTER GROUP foo DROP USER bar;
RTFM :)
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
On Mon, 19 Jun 2000, Nicolas Huillard wrote:
> I think all that works in 7.0, not 6.5 ?
Correct.
--
Peter Eisentraut Sernanders väg 10:115
[EMAIL PROTECTED] 75262 Uppsala
http://yi.org/peter-e/Sweden
gt; their privileges on it.
That could be a problem. Assume you delete all users from that table. How
do you get back in and add any? But something like this is clearly needed,
it just can't be in the database -- probably pg_hba.conf.
--
Peter Eisentraut
PGresult* res2 = PQexec (...);
> // Process res2.
> PQclear (res2);
> // return back to res1
> }
res1 and res2 are completely independent, you can use both as long and
often as you want. Not sure how you want to "include" one in the other
though.
--
Peter Eisentraut
mplete.o(.text+0x74): undefined reference to
> `rl_completion_append_character'
> tab-complete.o(.text+0x2654): undefined reference to
> `rl_completion_append_character'
> tab-complete.o(.text+0x2658): undefined reference to
> `rl_completion_append_character'
> gmake: *** [ps
't see the old table.
There's no way to "plug-in" tables files like that. You are more or less
out of luck. (The reason is that the table file only contains half the
truth. The other half is in the pg_log file that used to belong to it, but
it seems you don
; plperl.c:327: (Each undeclared identifier is reported only
> once
> plperl.c:327: for each function it appears in.)
> plperl.c:333: 'na' undeclared (first use in this function)
>
> Has anyone seen this before? I can fix the C code if I can find out
> w
ux system to access PG as
> root.
Perhaps you are confused. There is nothing wrong with having a database
user called "root". This user does not automatically have special
privileges and is unrelated to any Unix root user.
> Is it possible to mix two auth modes ?
Not from
Dronamraju Rajesh writes:
> /bin/sh: lorder: not found
> /bin/sh: tsort: not found
> /bin/sh: ar: not found
It seems you need to install a few more things. Not sure what Sun calls
this package, perhaps "binary utilities" or "compiler support".
--
Peter Eis
ting pg_log is almost like deleting the
data itself. You make backups, don't you?
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
which is inferred what tuples are valid. So
theoretically you could set "all transactions have committed" in pg_log
and pick out the tuples you like from the tables (after having gotten past
the horribly corrupted indexes). But that seems like a pretty complicated
undertaking in any case.
-
tbl_voters_full
> -rw---1 postgres postgres 144687104 Oct 5 00:09 tbl_voters_full.1
The table "tbl_voters" exceeded 1 GB in size and was thus split up into
two physical files. You will get .2, .3, etc. each time you grow another
GB.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
tmaster running at 'localhost'
> and accepting connections on Unix socket '5432'?
Check that there's a socket file /tmp/.s.PSQL.5432. Maybe it got deleted
accidentally. Restart the postmaster to create a new one.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
Loïc TREGOUËT writes:
> insert into tempo select id, name from topics order by name;
> ERROR: ORDER BY is not allowed in INSERT/SELECT
What's the point of this command? The values in "tempo" are going to end
up in random order anyway.
--
Peter Eisentraut [EMAIL
R D writes:
> In the TODO list is mentioned that in PG-7.1 you will have "More
> access control over who can create tables and use locks".
The TODO list is wrong.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
bly failed for you. You can read the file config.log to check
why. I'd be interested to know.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
ommand to find the
name.
> ps. list archives are unfortunately unusable...
Use www.geocrawler.com.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
Marcin Mazurek writes:
> I'm trying to give a createuser privilige by changing pg_users, i'm logged
> in as a superuser.
You need to change pg_shadow.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
s this the right mailing list to discuss this issue?
Generally, we discuss development on pgsql-hackers.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
Define "did not work".
though this way it is working fine on my local intranet
> which is on Linux.
What is "this way" and what is the nature of your intranet?
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
Jens Hartwig writes:
> is there any way to identify users which are currently connected to the
> database? Does perhaps even exist a method to notify this users (e.g. to
> close the connection)?
ps ax | grep postgres
is the best way to go.
--
Peter Eisentraut [EMAIL
out the identity of
the person that connected to your web page?
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
/docs/postgres/reference.htm
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
(
You're gonna have to, after which you'll immediately encounter linking
problems because the ODBC driver only compiles with GCC. Fixing that will
require some non-trivial programming work. Sorry.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
Williams Pedrogan writes:
> Please. PostgreSQL is installable on WIN98/M3?
Nope
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
write user
> in 'ps' command.
> I'm use FreeBSD 4.2.
Doesn't work on FreeBSD 4.* in PostgreSQL 7.0.*. Sorry.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
Showing actual error messages would be a reasonable start from your side.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
> The problem might be the following:
We are really not interested in what the problem "might be". As long as
you don't show us the error message as it appears on the screen you won't
get any help.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
the
> todo list of 7.1 but now it is removed. Stupid...
It was not possible to implement it without busting pg_dumpall, thus
creating an upgrading nightmare. It will be in 7.2.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
ed users,
so it has to be evaluated carefully.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
st process is the "postmaster" you started (by init
apparently), the second and third ones got forked to handle connections.
(one process per connection)
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
information.
> so you'll end up with multiple processes names 'postmaster', and have
> to track down which is which they way Peter pointed out.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
the user as which the server runs does not have to be named
"postgres".
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
the IPC sys V blablah then.
> I know N is the number of backend server processes, but does it have a
> result on the maximum number of clients ?
One backend server process per client.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
do things a user
> normally could not do. But given that I could create a rule on a
> pg_tables, one wonders if such a rule could be used to keep users from
> making tables.
It couldn't, because the CREATE TABLE code does not go through the rule
system.
--
Peter Eisentraut
4, .. yet it seemed to be requiring ODBC by default??
>
> 2) I don't know if this was related to #1, but none of the utilities
> installed (i.e. initdb), so I am dead in the water.
Show what you did and what appeared on the screen, not descriptions in
your own words.
--
Peter Eisen
to follow the FAQ to initialize the database structure, ..
> but found that there was no 'initdb' in the bin directory.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
aces/libpq all|install'.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
e". I already tried
> ftp://ftp.gnu.org. but I wasn't able to find it!
Try: ftp://ftp.sunfreeware.com/pub/freeware/sparc/2.6/make-3.76.1-sol26-sparc-local.gz
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
[EMAIL PROTECTED] writes:
> What showed up was the "Error index_formtuple: data takes 21268 bytes: too
> big". If anyone has any ideas on this, please share them.
It means your data is too big to fit into an index.
--
Peter Eisentraut [EMAIL PROTECTED] http://yi.org/peter-e/
1 - 100 of 508 matches
Mail list logo
