[BUGS] BUG #1493: image storeg
The following bug has been logged online: Bug reference: 1493 Logged by: naresh soni Email address: [EMAIL PROTECTED] PostgreSQL version: 7.4.1 Operating system: window 2000 professional Description:image storeg Details: how to store the images in the posgreSQL Manager,and how to edit the image ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [BUGS] BUG #1487: Index problem
Yes, it really depends on Locale. If I created a database with Locale = C, the problem won't happen (at least so far). BTW, I forgot to mention I tested with some Chinese and Japanese characters. However, it raises another issue why a wrong Locale would damage index? It shall only affect records with wrong conversion, not something like '1000'. The real environment in my program is J2EE+JDBC. I simply used pgAdmin to demonstrate the problem (not caused by JDBC or so). Though Locale C solved the problem, LIKE 'prefix%' still uses Seq Scan. It is a job to change hundreds (if not thousands SQL) to use BETWEEN/AND instead. Any suggestion? -Original Message- From: Dave Page [mailto:[EMAIL PROTECTED] Sent: Tuesday, February 22, 2005 12:19 AM To: Richard Huxton Cc: Tom Yeh; pgsql-bugs@postgresql.org Subject: RE: [BUGS] BUG #1487: Index problem > -Original Message- > From: Richard Huxton [mailto:[EMAIL PROTECTED] > Sent: 21 February 2005 16:13 > To: Dave Page > Cc: Tom Yeh; pgsql-bugs@postgresql.org > Subject: Re: [BUGS] BUG #1487: Index problem > > > pgAdmin uses libpq, just as psql does, so I cannot imagine why this > > would be the case. pgAdmin also does not do anything to try > to affect > > query plans in any way. > > Could it set the encoding/locale differently to psql (on the > same machine)? Good point, yes - it will try to set the encoding to unicode if it can. /D __ Do you Yahoo!? The all-new My Yahoo! - Get yours free! http://my.yahoo.com ---(end of broadcast)--- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to [EMAIL PROTECTED] so that your message can get through to the mailing list cleanly
Re: [BUGS] problem installing postgresql-8.0.1 on Windows 2K
Magnus, > You need to look a bit earlier in the logfile to determine where the > error is - the log you have here is only from the MSI rollback. > > How far into the installation do you get befoer it fails? I get to the point where it would start copying the files (it shows the progressbar for a very very brief moment and then the error pops up). Anyway, I'm pasting the earlier lines where I think the fatal problem first appears (previously it complains about error 2769 on locales and 2888 and 2826 on TextStyle - but those I think are non-fatal). HAS_DOTNET="C:\WINNT\Microsoft.NET\Framework\" INTERNALLAUNCH="1" ALLUSERS="1" SERVICEOK="1" INITDBOK="1 " TARGETDIR="C:\" BIN *** Log Line Missing *** MSI (s) (90:50): Resetting cached policy values MSI (s) (90:50): Machine policy value 'Debug' is 0 MSI (s) (90:50): *** RunEngine: *** Product: C:\tmp\postgresql-8.0.1\postgresql-8.0-int.msi *** Action: INSTALL *** CommandLine: ** MSI (s) (90:50): Machine policy value 'DisableUserInstalls' is 0 MSI (s) (90:50): Setting cached product context: User non-assigned for product: 207809E353FA1164598159D52AB4E770 MSI (s) (90:50): Using cached product context: User non-assigned for product: 207809E353FA1164598159D52AB4E770 MSI (s) (90:50): Setting cached product context: User non-assigned for product: F12A43F7BED28954BB9116D1B62D2417 MSI (s) (90:50): Using cached product context: User non-assigned for product: F12A43F7BED28954BB9116D1B62D2417 MSI (s) (90:50): Setting cached product context: machine assigned for product: 017DB44EA17B3d11F9970006808AE88C *snip similar lines* MSI (s) (90:50): Setting cached product context: machine assigned for product: FB5C4BDE4233F014E8B106AA5B68C83C MSI (s) (90:50): Using cached product context: machine assigned for product: FB5C4BDE4233F014E8B106AA5B68C83C MSI (s) (90:50): Note: 1: 2318 2: C:\tmp\postgresql-8.0.1\postgresql-8.0-int.msi MSI (s) (90:50): MainEngineThread is returning 1603 MSI (c) (B8:70): Back from server. Return value: 1603 MSI (c) (B8:70): Decrementing counter to disable shutdown. If counter >= 0, shutdown will be denied. Counter after decrement: -1 Action ended 15:56:09: ExecuteAction. Return value 3. MSI (c) (B8:70): Doing action: Fatal_Error Action start 15:56:09: Fatal_Error. Action 15:56:09: Fatal_Error. Dialog created Action ended 15:56:11: Fatal_Error. Return value 2. Action ended 15:56:11: INSTALL. Return value 3. MSI (c) (B8:70): Destroying RemoteAPI object. Property(C): BASEDIR = C:\java\PostgreSQL\8.0\ Property(C): SERVERDIR = C:\java\PostgreSQL\8.0\ ... - Best regards, Juu. __ Advertisement: » LCD monitori saakot no 151 Ls » Pieziimju datori saakot no 527 Ls » Peerkot NOKIA vai SONY ERICSON mobilos telefonus, sanjem daavanaa aksesuaarus.
[BUGS] BUG #1497: Default permissions allow any user to create objects in any database
The following bug has been logged online: Bug reference: 1497 Logged by: Barry Brown Email address: [EMAIL PROTECTED] PostgreSQL version: 8.0.1 Operating system: RHEL 3.0 Description:Default permissions allow any user to create objects in any database Details: The docs say that initially only the owner of a database may use the objects created in it. But I have found that ANY user can work with any object by default, even in the template1 database. To reproduce the bug, I will create a user called 'sqltest' with password 'foobar', connect to template1 as that user, and create a table. First, create the user: [EMAIL PROTECTED] data]# su - postgres -bash-2.05b$ psql template1 Welcome to psql 8.0.1, the PostgreSQL interactive terminal. Type: \copyright for distribution terms \h for help with SQL commands \g or terminate with semicolon to execute query \q to quit template1=# create user sqltest password 'foobar'; CREATE USER Next, connect as that user to template1: [EMAIL PROTECTED] barry]$ psql -U sqltest template1 Password: Welcome to psql 8.0.1, the PostgreSQL interactive terminal. Type: \copyright for distribution terms \h for help with SQL commands \? for help with psql commands \g or terminate with semicolon to execute query \q to quit template1=> Finally, create a table: template1=> create table foo ( template1(> id_num int template1(> ); CREATE TABLE template1=> \dt foo List of relations Schema | Name | Type | Owner +--+---+- public | foo | table | sqltest (1 row) I should not be able to do this. I compiled PostgreSQL from the FC3 SRPM found on rpmfind.net. I had to disable SSL, KRB, and Selftest in the spec file to make it compile. ---(end of broadcast)--- TIP 8: explain analyze is your friend
Re: [BUGS] BUG #1491: Getting an "undefined symbol" when loading
Tom, Ha! If I'd known of the select version(); I would have fixed the problem sooner! It all works when you run the right version... 8-) You may want to specify that select command in the doc in regard to modules so people can test their backend version before to bug you guys. Thank you for the hint, very helpful. This ltree stuff is really cool! Alexis On Mon, 21 Feb 2005, Tom Lane wrote: > "Alexis Wilke" <[EMAIL PROTECTED]> writes: > > Operating system: Linux > > You need to be a lot more specific than that, because this stuff works > fine on all the variants of Linux we know about. > > > ERROR: Load of file /usr/local/pgsql/lib/ltree.so failed: > > /usr/local/pgsql/lib/ltree.so: undefined symbol: errstart > > Are you sure you're loading into a 7.4-or-later backend? (Check "select > version();" to be certain.) The only other possibility I can think of > is that the backend has been stripped of symbols --- in which case it's > not errstart that's at fault particularly, that just happens to be the > first missing symbol reported. > > regards, tom lane > ---(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
[BUGS] BUG #1496: Problem with instalation
The following bug has been logged online: Bug reference: 1496 Logged by: Sergius Email address: [EMAIL PROTECTED] PostgreSQL version: 8.0 Operating system: Win NT Description:Problem with instalation Details: With I make the instalation using the windows NT?? ---(end of broadcast)--- TIP 8: explain analyze is your friend
[BUGS] BUG #1498: Default permissions allow any user to create objects in any database (revised)
The following bug has been logged online: Bug reference: 1498 Logged by: Barry Brown Email address: [EMAIL PROTECTED] PostgreSQL version: 8.0.1 Operating system: RHEL 3.0 (kernel 2.4.21-27.0.2.ELsmp) Description:Default permissions allow any user to create objects in any database (revised) Details: I just submitted bug 1497. I did some more tests and have made it easier to reproduce the bug by compiling from scratch instead of using an SRPM. I have included all steps, starting from compilation. Building: postgres user is UID 26, GID 26 Unpacked -base and -opt. % ./configure --prefix=/usr/local % make # make install # mkdir /var/lib/pgsql # chown postgres:postgres /var/lib/pgsql # chmod 700 /var/lib/pgsql Changed the last lines of pg_hba.conf to this: # "local" is for Unix domain socket connections only #local all all ident sameuser # IPv4 local connections: #hostall all 127.0.0.1/32 ident sameuser # IPv6 local connections: #hostall all ::1/128 ident sameuser local template1postgres ident sameuser local all all md5 Starting: I have an /etc/init.d/postgresql script left over from the FC3 build. I changed PCENGINE to point to /usr/local/bin (instead of /usr/bin) so it will run the correct binaries. PGDATA is /var/lib/pgsql As root: [EMAIL PROTECTED] postgresql-8.0.1]# /etc/init.d/postgresql.local start Initializing database: [ OK ] Starting postgresql.local service: [ OK ] Create users and databases: [EMAIL PROTECTED] postgresql-8.0.1]# su - postgres -bash-2.05b$ /usr/local/bin/psql template1 Welcome to psql 8.0.1, the PostgreSQL interactive terminal. Type: \copyright for distribution terms \h for help with SQL commands \? for help with psql commands \g or terminate with semicolon to execute query \q to quit template1=# create user sqltest1 password 'foobar1'; CREATE USER template1=# create user sqltest2 password 'foobar2'; CREATE USER template1=# create database sqltest1db owner sqltest1; CREATE DATABASE Create a table in someone else's database. First, connect as user sqltest2 to the sqltest1db database: [EMAIL PROTECTED] barry]$ /usr/local/bin/psql -U sqltest2 sqltest1db Password: Welcome to psql 8.0.1, the PostgreSQL interactive terminal. Type: \copyright for distribution terms \h for help with SQL commands \? for help with psql commands \g or terminate with semicolon to execute query \q to quit sqltest1db=> Then create a table: sqltest1db=> create table foo ( id int ); CREATE TABLE sqltest1db=> \dt foo List of relations Schema | Name | Type | Owner +--+---+-- public | foo | table | sqltest2 (1 row) This is the bug. The user sqltest2 should not be able to create a table in sqltest1's database with being given permissions to do so. Worse yet, any valid database user can connect to the template1 database and create tables in it. ---(end of broadcast)--- TIP 2: you can get off all lists at once with the unregister command (send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])
Re: [BUGS] BUG #1490: contrib .sql.in sed command bogus
Hi Tom, Indeed, with the correct version the $libdir will work in the .sql file. Very good. 8-) Alexis On Mon, 21 Feb 2005, Tom Lane wrote: > "Alexis Wilke" <[EMAIL PROTECTED]> writes: > > Inside pgxs.mk I had to change the $$libdir into $(libdir ) to get the > > proper path in the final .sql file: > > > > ifneq (,$(MODULES)$(MODULE_big)) > > %.sql: %.sql.in > > sed 's,MODULE_PATHNAME,$(libdir)/$*,g' $< >$@ > > endif > > This is not correct --- the substitution is supposed to be literally > "$libdir", and so the makefile is correct as-is. (I'm not sure why we > still bother with doing the sed conversion at all, except that if we > removed it you can be sure we'd think of some other thing we needed it > for ;-)) > > regards, tom lane > ---(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
[BUGS] BUG #1494: psql \df to_char
The following bug has been logged online: Bug reference: 1494 Logged by: Ludwig Lim Email address: [EMAIL PROTECTED] PostgreSQL version: 8.0.1 Operating system: Linux Description:psql \df to_char Details: Hi: I notice that when I type \df to_char in psql prompt the following does not appear as one of the possible arguments of to_char: time without time zone,text However, the following select statement works: SELECT to_char(current_time :: 'time without time zone', 'ss'); Is this a bug? Version of PostgreSQL: PostgreSQL 8.0.1 on i686-pc-linux-gnu compiled by GCC 2.96 Thanks in advance, Ludwig Lim ---(end of broadcast)--- TIP 7: don't forget to increase your free space map settings
[BUGS] BUG #1504: Wrong user is used for sequences through rules
The following bug has been logged online:
Bug reference: 1504
Logged by: Nicolas Rachinsky
Email address: [EMAIL PROTECTED]
PostgreSQL version: 7.4.7
Operating system: FreeBSD 4.10
Description:Wrong user is used for sequences through rules
Details:
As the user (nicolas) owning the database (webtempl) I execute:
DROP TABLE log CASCADE;
CREATE TABLE log(
id BIGSERIAL PRIMARY KEY,
msg TEXT NOT NULL DEFAULT ''
);
DROP TABLE log_insert CASCADE;
CREATE TABLE log_insert(
msg TEXT NOT NULL DEFAULT ''
);
CREATE RULE log_ins AS ON INSERT TO log_insert
DO INSTEAD
INSERT INTO log (msg) VALUES (NEW.msg);
GRANT insert
ON log_insert
TO web;
snip
Now as user 'web' I try the following:
webtempl=> insert into log_insert (msg) VALUES ('foo');
ERROR: permission denied for sequence log_id_seq
webtempl=>
It works after executing the following as the user owning the database:
GRANT update
ON log_id_seq
TO web;
---snip---
As user 'web':
webtempl=> insert into log_insert (msg) VALUES ('foo');
INSERT 1078083 1
webtempl=>
As I read
http://www.postgresql.org/docs/7.4/interactive/rules-privileges.html the
insert should
work without granting the update privilege.
version
--
PostgreSQL 7.4.7 on i386-portbld-freebsd4.10, compiled by GCC 2.95.4
Thanks,
Nicolas
---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])
[BUGS] BUG #1503: pg_restore cannot restore a backup
The following bug has been logged online: Bug reference: 1503 Logged by: Yavuz Email address: [EMAIL PROTECTED] PostgreSQL version: 8.0.1 Operating system: windows 2000 sp4 Description:pg_restore cannot restore a backup Details: firstly i try a backup on Fedore Linux core2 with 8.0.1 pg_dump -f backup.tar -Ft -U postgres mydb and and restore it with createdb mydb2 -U postgres pg_restore -d mydb2 backup.tar -U postgres it works very fine and i try this commands on win2000 service pack4 with postgresql 8.0.1 pg_dump -f backup.tar -Ft -U postgres mydb works well and i can see the listing with pg_restore -l backup.tar and a create a database to restore with createdb mydb2 -U postgres its ok but when i try pg_restore -d mydb2 backup.tar -U postgres i have an error like "pg_restore:[archiver(db)] connection to database "mydb2" failed: FATAL : password authentication failed for user "postgres"" i check my password on psql , pg_dump and other programs there is no problem ---(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
[BUGS] BUG #1499: Unable to install
The following bug has been logged online: Bug reference: 1499 Logged by: Luke Lincoln Email address: [EMAIL PROTECTED] PostgreSQL version: None Yet! Operating system: Windows 2000 Description:Unable to install Details: Error message (using Either msi file) "Please use the main MSI file to install PostgreSQL" Thank you! ---(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
[BUGS] BUG #1500: child dead
The following bug has been logged online: Bug reference: 1500 Logged by: nagash Email address: [EMAIL PROTECTED] PostgreSQL version: 8.0.1 Operating system: RedHat Description:child dead Details: site_utf=# SELECT to_char(NOW() - '2000-10-02'::date, 'MonDD'); server closed the connection unexpectedly This probably means the server terminated abnormally before or while processing the request. The connection to the server was lost. Attempting reset: Failed. tested on 2 servers and it show with exactly 'MonDD' if I write SELECT to_char(NOW() - '2000-10-02'::date, ''); it will be OK ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
Re: [BUGS] BUG #1497: Default permissions allow any user to create objects in any database
"Barry Brown" <[EMAIL PROTECTED]> writes: > I should not be able to do this. Not everyone would agree with you. But if you want to work that way, revoke public CREATE privileges on the public schema, and revoke public CREATE privileges on the database so people can't create new schemas. regards, tom lane ---(end of broadcast)--- TIP 4: Don't 'kill -9' the postmaster
[BUGS] BUG #1502: hash_seq_search might return removed entry
The following bug has been logged online: Bug reference: 1502 Logged by: Thomas Email address: [EMAIL PROTECTED] PostgreSQL version: 8.0.1 Operating system: N/A Description:hash_seq_search might return removed entry Details: The hash_seq_search keeps track of what element that it should return next in a HASH_SEQ_STATUS struct when it peruses a bucket. Removing that element from the table won't change anything since the struct remains unaffected. It still holds onto that element and hence, will return it on next iteration. ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq
[BUGS] BUG #1501: in additional to previous "DEAD of child"
The following bug has been logged online: Bug reference: 1501 Logged by: nagash Email address: [EMAIL PROTECTED] PostgreSQL version: 7.4.2 Operating system: gentoo Description:in additional to previous "DEAD of child" Details: pg kill all it's childs also tested on 7.4.2 PG on gentoo OS as result: the same bug ---(end of broadcast)--- TIP 5: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq
