[BUGS] packaging for AIX - compliments!!
I would like to compliment all of you on how clean the packaging for AIX is
going for me. (I still have to test that they actually work though :))
Using IBM compiler and this configure command + make - I am getting through
the make process.
I have to check whether your make also uses the GNU convention for
performing an install in DESTDIR - as I use that to create installable
packages in AIX installp format.
CFLAGS="-O2 -qcpluscmt" ./configure \
--prefix=/opt \
--sysconfdir=/var/postgresql/etc \
--sharedstatedir=/var/postgresql/com \
--localstatedir=/var/postgresql \
--mandir=/usr/share/man --without-readline
Yesterday and today - with no noteable issues I have packaged from:
*postgresql-7.4.30.tar.bz2
postgresql-8.0.26.tar.bz2
postgresql-8.1.23.tar.bz2
postgresql-8.2.23.tar.bz2* (in progress)
Note: there are lots of warning messages such as:
"pg_some_file.c", line 402.33: 1506-068 (W) Operation between types "struct
{...}*" and "struct gzFile_s*" is not allowed.
*postgresql-8.3.22.tar.bz2* (in progress)
Unpacking the archive has an "error". There are couple references to a file
@LongLink (contents)
$ cat @Long*
postgresql-8.3.22/src/backend/utils/mb/conversion_procs/utf8_and_euc_jis_2004/utf8_and_euc_jis_2004.c
Which should not be in the archive - I think.
I shall continue with the rest to see how far they come - and update this
thread with those results.
Again, thanks for the clean builds - and Happy Holidays!
Re: [BUGS] BUG #7760: records with duplicate primary keys
On 20/12/2012 1:22 AM, acummi...@aperiogroup.com wrote: > The following bug has been logged on the website: > > Bug reference: 7760 > Logged by: Alan Cummings > Email address: acummi...@aperiogroup.com > PostgreSQL version: 9.1.7 > Operating system: Ubuntu Linux 10.04.1 > Description: > > We found 6 duplicated primary keys in a table. All fields were identical in > the duplicated records, including the primary key. (Total records ~ 22,000). > We recently updated from 9.1.2 to 9.1.7, so the issue may have been present > in 9.1.2. Have you had your database running with fsync=off at any point? What storage system is the database on? RAID array? If so, details - drive types, controller if any, writeback/writethrough, BBU enabled,etc? SSD, and if so which SSD? Has the DB host unexpectedly powered off or rebooted at any point? This should never cause a problem unless you run with unsafe settings like fsync=off, but is important to know. Did you REINDEX as part of the upgrade to 9.1.7? A corrupt unique index could hide duplicate records, and a reindex might well cause the problem to become apparent. > We have fixed the issue in our table by backing up the table in pgAdmin to a > plain file, deleting the duplicated records in the text file, then restoring > the table. > > I can send the original backup with the duplicated records if that would be > helpful -- it is about 11MB. Is the backup a `pg_dump`? Or a raw copy of the database directory? The pg_dump output won't usually provide much information, it's the actual database on disk that's significant. -- Craig Ringer http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Training & Services -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
[BUGS] BUG #7771: packaging error causes build failure
The following bug has been logged on the website: Bug reference: 7771 Logged by: Michael Felt Email address: mamf...@gmail.com PostgreSQL version: 8.3.22 Operating system: AIX Description: There are three references to the file @LongLink in the tar file : postgresql-8.3.22.tar.bz2 michael@x054:[/data/prj/pgsql]bzip2 -dc postgresql-8.3.22.tar.bz2 | tar tf - | grep LongLink ././@LongLink ././@LongLink ././@LongLink $ cat @LongLink postgresql-8.3.22/src/backend/utils/mb/conversion_procs/utf8_and_euc_jis_2004/utf8_and_euc_jis_2004.c make stops with following message. Looks as if the source file referenced in @LongLink needed is missing 356 1500-010: (W) WARNING in BackgroundWriterMain: Infinite loop. Program may not stop. 221 1500-010: (W) WARNING in WalWriterMain: Infinite loop. Program may not stop. 3638 1500-010: (W) WARNING in PostgresMain: Infinite loop. Program may not stop. 1500-030: (I) INFORMATION: get_rule_expr: Additional optimization may be attained by recompiling and specifying MAXMEM option with a value greater than 8192. 1500-030: (I) INFORMATION: dch_date: Additional optimization may be attained by recompiling and specifying MAXMEM option with a value greater than 8192. make[3]: *** No rule to make target `utf8_and_euc_jis_2004.o', needed by `libutf8_and_euc_jis_2004.so.0'. Stop. make[2]: *** [all] Error 2 make[1]: *** [all] Error 2 make: *** [all] Error 2 + make install DESTDIR=/var/tmp/root/postgresql-8.3.22.0 > build/aix/install.out make[3]: *** No rule to make target `utf8_and_euc_jis_2004.o', needed by `libutf8_and_euc_jis_2004.so.0'. Stop. make[2]: *** [install] Error 2 make[1]: *** [install] Error 2 make: *** [install] Error 2 make install returned an error michael@x054:[/data/prj/pgsql]find . -name utf8_and_euc_jis_2004.c michael@x054:[/data/prj/pgsql] p.s. @LongLink is also in the older distribution .bz2 files, I had just not had the time to look into the error it caused. I hope you can at least re-release this tar file so that all the files needed are in it! -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] BUG #7771: packaging error causes build failure
File postgresql-8.4.15.tar.bz2 has the same error as postgresql-8.3.22.tar.bz2 michael@x054:[/data/prj/pgsql]bzip2 -dc postgresql-8.3.22.tar.bz2 | tar tf - | grep LongLink ././@LongLink ././@LongLink ././@LongLink michael@x054:[/data/prj/pgsql]find . -name utf8_and_euc_jis_2004.c michael@x054:[/data/prj/pgsql]rm @LongLink rm: Remove @LongLink? y michael@x054:[/data/prj/pgsql]bzip2 -dc postgresql-8.4.15.tar.bz2 | tar xf - tar: 0511-188 Cannot create ././@LongLink: The file access permissions do not allow the specified action. tar: 0511-188 Cannot create ././@LongLink: The file access permissions do not allow the specified action. michael@x054:[/data/prj/pgsql]chmod a+r @LongLink michael@x054:[/data/prj/pgsql]cat @LongLink postgresql-8.4.15/src/backend/utils/mb/conversion_procs/utf8_and_euc_jis_2004/utf8_and_euc_jis_2004.c = The version 9 files are clean of this issue: michael@x054:[/data/prj/pgsql]bzip2 -dc postgresql-9.0.11.tar.bz2 | tar xf - michael@x054:[/data/prj/pgsql]bzip2 -dc postgresql-9.1.7.tar.bz2 | tar xf - michael@x054:[/data/prj/pgsql]bzip2 -dc postgresql-9.2.2.tar.bz2 | tar xf - = On Mon, Dec 24, 2012 at 11:30 AM, wrote: > The following bug has been logged on the website: > > Bug reference: 7771 > Logged by: Michael Felt > Email address: mamf...@gmail.com > PostgreSQL version: 8.3.22 > Operating system: AIX > Description: > > There are three references to the file @LongLink in the tar file : > postgresql-8.3.22.tar.bz2 > > michael@x054:[/data/prj/pgsql]bzip2 -dc postgresql-8.3.22.tar.bz2 | tar > tf - > | grep LongLink > ././@LongLink > ././@LongLink > ././@LongLink > > > $ cat @LongLink > > postgresql-8.3.22/src/backend/utils/mb/conversion_procs/utf8_and_euc_jis_2004/utf8_and_euc_jis_2004.c > > make stops with following message. Looks as if the source file referenced > in > @LongLink needed is missing > > 356 1500-010: (W) WARNING in BackgroundWriterMain: Infinite loop. > Program may not stop. > 221 1500-010: (W) WARNING in WalWriterMain: Infinite loop. Program > may not stop. > 3638 1500-010: (W) WARNING in PostgresMain: Infinite loop. Program > may > not stop. > 1500-030: (I) INFORMATION: get_rule_expr: Additional optimization may > be > attained by recompiling and specifying MAXMEM option with a value greater > than 8192. > 1500-030: (I) INFORMATION: dch_date: Additional optimization may be > attained by recompiling and specifying MAXMEM option with a value greater > than 8192. > make[3]: *** No rule to make target `utf8_and_euc_jis_2004.o', needed by > `libutf8_and_euc_jis_2004.so.0'. Stop. > make[2]: *** [all] Error 2 > make[1]: *** [all] Error 2 > make: *** [all] Error 2 > + make install DESTDIR=/var/tmp/root/postgresql-8.3.22.0 > > build/aix/install.out > make[3]: *** No rule to make target `utf8_and_euc_jis_2004.o', needed by > `libutf8_and_euc_jis_2004.so.0'. Stop. > make[2]: *** [install] Error 2 > make[1]: *** [install] Error 2 > make: *** [install] Error 2 > make install returned an error > > michael@x054:[/data/prj/pgsql]find . -name utf8_and_euc_jis_2004.c > michael@x054:[/data/prj/pgsql] > > p.s. @LongLink is also in the older distribution .bz2 files, I had just not > had the time to look into the error it caused. I hope you can at least > re-release this tar file so that all the files needed are in it! > > > > > -- > Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) > To make changes to your subscription: > http://www.postgresql.org/mailpref/pgsql-bugs >
Re: [BUGS] packaging for AIX - compliments!!
See bug #7771 for packaging issue preventing build for 8.3.22 and 8.4.15.
Will start with the version 9.X.y now
On Mon, Dec 24, 2012 at 11:17 AM, Michael Felt wrote:
> I would like to compliment all of you on how clean the packaging for AIX
> is going for me. (I still have to test that they actually work though :))
>
> Using IBM compiler and this configure command + make - I am getting
> through the make process.
> I have to check whether your make also uses the GNU convention for
> performing an install in DESTDIR - as I use that to create installable
> packages in AIX installp format.
>
> CFLAGS="-O2 -qcpluscmt" ./configure \
> --prefix=/opt \
> --sysconfdir=/var/postgresql/etc \
> --sharedstatedir=/var/postgresql/com \
> --localstatedir=/var/postgresql \
> --mandir=/usr/share/man --without-readline
>
>
> Yesterday and today - with no noteable issues I have packaged from:
> *postgresql-7.4.30.tar.bz2
> postgresql-8.0.26.tar.bz2
> postgresql-8.1.23.tar.bz2
> postgresql-8.2.23.tar.bz2* (in progress)
>
> Note: there are lots of warning messages such as:
> "pg_some_file.c", line 402.33: 1506-068 (W) Operation between types
> "struct {...}*" and "struct gzFile_s*" is not allowed.
>
> *postgresql-8.3.22.tar.bz2* (in progress)
> Unpacking the archive has an "error". There are couple references to a
> file @LongLink (contents)
> $ cat @Long*
>
> postgresql-8.3.22/src/backend/utils/mb/conversion_procs/utf8_and_euc_jis_2004/utf8_and_euc_jis_2004.c
>
> Which should not be in the archive - I think.
>
> I shall continue with the rest to see how far they come - and update this
> thread with those results.
>
> Again, thanks for the clean builds - and Happy Holidays!
>
Re: [BUGS] packaging for AIX - compliments!!
On 12/24/2012 2:17 AM, Michael Felt wrote: *postgresql-8.3.22.tar.bz2* (in progress) Unpacking the archive has an "error". There are couple references to a file @LongLink (contents) $ cat @Long* postgresql-8.3.22/src/backend/utils/mb/conversion_procs/utf8_and_euc_jis_2004/utf8_and_euc_jis_2004.c Which should not be in the archive - I think. try using gtar, from the IBM Linux toolkit. I'm pretty sure 8.3 built fine for me on AIX, but recently I've only been building 8.4, and 9.0,9.1
Re: [BUGS] BUG #7771: packaging error causes build failure
mamf...@gmail.com writes: > There are three references to the file @LongLink in the tar file : > postgresql-8.3.22.tar.bz2 > michael@x054:[/data/prj/pgsql]bzip2 -dc postgresql-8.3.22.tar.bz2 | tar tf - > | grep LongLink > ././@LongLink > ././@LongLink > ././@LongLink No, there is no such thing in our tarfiles. Apparently you have a version of tar that's been hacked to print that if it sees a pathname too long for it to cope with. In 9.0 and up we renamed a couple of deeply-nested files to shorten the maximum pathlength in the tarballs, cf http://archives.postgresql.org/pgsql-general/2009-11/msg00078.php http://git.postgresql.org/gitweb/?p=postgresql.git&a=commitdiff&h=6bef82b38a67b493fc82990e41946c1bd2ea1b03 If you can't readily find a less broken version of tar for your platform, you might be best advised to just forget about the 8.x releases. I'm not sure why you'd bother with them anyway, still less with branches that are long out of support. regards, tom lane -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
[BUGS] "text to text" operator redefinition ignored
PG 9.2.2, Windows, empty database just curious, redefinition of various operators seems to work nicely but "text to text" operator redefinition (last block down here) seems completely ignored: set search_path to public; select 'aa'::varchar = 'aa '::varchar; -- returns false create or replace function public.opr_vceqvc(varchar, varchar) returns boolean as $$ select rtrim($1)=rtrim($2) $$ language sql immutable; create operator public.= (leftarg = varchar, rightarg = varchar, procedure = public.opr_vceqvc, commutator = =, negator = <>, restrict = eqsel, join = eqjoinsel, hashes, merges); select 'aa'::varchar = 'aa '::varchar; -- returns true drop operator if exists public.= (varchar, varchar); select 'aa'::text = 'aa '::varchar; -- returns false create or replace function public.opr_txeqvc(text, varchar) returns boolean as $$ select rtrim($1)=rtrim($2) $$ language sql immutable; create operator public.= (leftarg = text, rightarg = varchar, procedure = public.opr_txeqvc, commutator = =, negator = <>, restrict = eqsel, join = eqjoinsel, hashes, merges); select 'aa'::text = 'aa '::varchar; -- returns true drop operator if exists public.= (text, varchar); select 'aa'::text = 'aa '::text; -- returns false create or replace function public.opr_txeqtx(text, text) returns boolean as $$ -- select rtrim($1)=rtrim($2); select true; $$ language sql immutable; create operator public.= (leftarg = text, rightarg = text, procedure = public.opr_txeqtx, commutator = =, negator = <>, restrict = eqsel, join = eqjoinsel, hashes, merges); select 'aa'::text = 'aa '::text; -- ALWAYS RETURNS FALSE! drop operator if exists public.= (text, text); -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] "text to text" operator redefinition ignored
t...@lavabit.com writes: > PG 9.2.2, Windows, empty database > just curious, redefinition of various operators seems to work nicely but > "text to text" operator redefinition (last block down here) seems > completely ignored: > set search_path to public; That search path setting doesn't do what you probably think. See under "The System Catalog Schema" in http://www.postgresql.org/docs/9.2/static/ddl-schemas.html Other relevant documentation is in http://www.postgresql.org/docs/9.2/static/typeconv-oper.html regards, tom lane -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
