How good is FreeBSD for postgres ?

Started by Oleg Bartunovover 26 years ago7 messages

oleg@sai.msu.su

over 26 years ago

Hi,

I have to work with postgres under freeBSD 3.1-RELEASE and as I'm
completely dummy in FreeBSD I'm wondering if there are problems
with running postgres (6.5).

I installed 6.5 cvs and after playing with kernel (config. shared memory)
I ran regression tests.There were much more tests failures than in
case of my lovely :-) Linux.
Is this a known problem with latest 6.5 cvs or it's my fault ?
I used gcc 2.7.2.1.

float8 .. failed
geometry .. failed
create_function_2 .. failed
triggers .. failed
sanity_check .. failed
misc .. failed
alter_table .. failed
rules .. failed
plpgsql .. failed

I attached below some diffs.

Regards,

Oleg

diff results/float8.out expected/float8.out

190,191d189
< ERROR: floating point exception! The last floating point operation either exc
eeded legal ranges or was a divide by zero
< QUERY: SELECT '' AS bad, f.f1 ^ '1e200' from FLOAT8_TBL f;
192a191,192

QUERY: SELECT '' AS bad, f.f1 ^ '1e200' from FLOAT8_TBL f;
ERROR: pow() result is out of range

198,206c198
< bad| ?column?
< ---+--------------------
< | 1
< |7.39912306090513e-16
< | 0
< | 0
< | 1
< (5 rows)
<
---

ERROR: exp() result is out of range

223a216

ERROR: Input '10e-400' is out of range for float8

224a218

ERROR: Input '-10e-400' is out of range for float8

diff results/create_function_2.out expected/create_function_2.out
41,44d40
< pqReadData() -- backend closed the channel unexpectedly.
<       This probably means the backend terminated abnormally
<       before or while processing the request.
< We have lost the connection to the backend, so further processing is impossibl
e.  Terminating.

diff results/triggers.out expected/triggers.out
39,42c39,285
< pqReadData() -- backend closed the channel unexpectedly.
<       This probably means the backend terminated abnormally
<       before or while processing the request.
< We have lost the connection to the backend, so further processing is impossibl
e.  Terminating.
---

QUERY: insert into fkeys2 values (30, '3', 2);

..... cutted .....

diff results/sanity_check.out expected/sanity_check.out
13,14d12
< fkeys         |t          
< fkeys2        |t          
29d26
< pkeys         |t          
34c31
< (26 rows)
---

(23 rows)

diff results/misc.out expected/misc.out
9,12c9,514
< pqReadData() -- backend closed the channel unexpectedly.
<       This probably means the backend terminated abnormally
<       before or while processing the request.
< We have lost the connection to the backend, so further processing is impossibl
e.  Terminating.
---

QUERY: UPDATE tmp

....... cutted .....

diff results/plpgsql.out expected/plpgsql.out
875d874
< ERROR:  Load of file /usr/local/pgsql/lib/plpgsql.so failed: dlopen '/usr/loca
l/pgsql/lib/plpgsql.so' failed. (/usr/local/pgsql/lib/plpgsql.so: Undefined symb
ol "CurrentMemoryContext")
877d875
< ERROR:  Can't find function plpgsql_call_handler in file /usr/local/pgsql/lib/
plpgsql.so
879d876
< ERROR:  Can't find function plpgsql_call_handler in file /usr/local/pgsql/lib/
plpgsql.so
.... cutted ...

_____________________________________________________________
Oleg Bartunov, sci.researcher, hostmaster of AstroNet,
Sternberg Astronomical Institute, Moscow University (Russia)
Internet: oleg@sai.msu.su, http://www.sai.msu.su/~megera/
phone: +007(095)939-16-83, +007(095)939-23-83

The Hermit Hacker

scrappy@hub.org

over 26 years ago

In reply to: Oleg Bartunov (#1)

Re: [HACKERS] How good is FreeBSD for postgres ?

All known problems...I've always run PostgreSQL under FreeBSD, and have
yet to notice a problem with a -release version of it...

The regression tests are "base Linux", and some of the generated error
messages are different between the systems and/or rounding is slighty
different...

On Sat, 15 May 1999, Oleg Bartunov wrote:

Hi,

I have to work with postgres under freeBSD 3.1-RELEASE and as I'm
completely dummy in FreeBSD I'm wondering if there are problems
with running postgres (6.5).

I installed 6.5 cvs and after playing with kernel (config. shared memory)
I ran regression tests.There were much more tests failures than in
case of my lovely :-) Linux.
Is this a known problem with latest 6.5 cvs or it's my fault ?
I used gcc 2.7.2.1.

float8 .. failed
geometry .. failed
create_function_2 .. failed
triggers .. failed
sanity_check .. failed
misc .. failed
alter_table .. failed
rules .. failed
plpgsql .. failed

I attached below some diffs.

Regards,

Oleg
diff results/float8.out expected/float8.out
190,191d189
< ERROR: floating point exception! The last floating point operation either exc
eeded legal ranges or was a divide by zero
< QUERY: SELECT '' AS bad, f.f1 ^ '1e200' from FLOAT8_TBL f;
192a191,192

QUERY: SELECT '' AS bad, f.f1 ^ '1e200' from FLOAT8_TBL f;
ERROR: pow() result is out of range

198,206c198
< bad| ?column?
< ---+--------------------
< | 1
< |7.39912306090513e-16
< | 0
< | 0
< | 1
< (5 rows)
<
---

ERROR: exp() result is out of range

223a216

ERROR: Input '10e-400' is out of range for float8

224a218

ERROR: Input '-10e-400' is out of range for float8
diff results/create_function_2.out expected/create_function_2.out
41,44d40
< pqReadData() -- backend closed the channel unexpectedly.
<       This probably means the backend terminated abnormally
<       before or while processing the request.
< We have lost the connection to the backend, so further processing is impossibl
e.  Terminating.
diff results/triggers.out expected/triggers.out
39,42c39,285
< pqReadData() -- backend closed the channel unexpectedly.
<       This probably means the backend terminated abnormally
<       before or while processing the request.
< We have lost the connection to the backend, so further processing is impossibl
e.  Terminating.
---
QUERY: insert into fkeys2 values (30, '3', 2);

..... cutted .....
diff results/sanity_check.out expected/sanity_check.out
13,14d12
< fkeys         |t          
< fkeys2        |t          
29d26
< pkeys         |t          
34c31
< (26 rows)
---
(23 rows)
diff results/misc.out expected/misc.out
9,12c9,514
< pqReadData() -- backend closed the channel unexpectedly.
<       This probably means the backend terminated abnormally
<       before or while processing the request.
< We have lost the connection to the backend, so further processing is impossibl
e.  Terminating.
---
QUERY: UPDATE tmp

....... cutted .....
diff results/plpgsql.out expected/plpgsql.out
875d874
< ERROR:  Load of file /usr/local/pgsql/lib/plpgsql.so failed: dlopen '/usr/loca
l/pgsql/lib/plpgsql.so' failed. (/usr/local/pgsql/lib/plpgsql.so: Undefined symb
ol "CurrentMemoryContext")
877d875
< ERROR:  Can't find function plpgsql_call_handler in file /usr/local/pgsql/lib/
plpgsql.so
879d876
< ERROR:  Can't find function plpgsql_call_handler in file /usr/local/pgsql/lib/
plpgsql.so
.... cutted ...
_____________________________________________________________
Oleg Bartunov, sci.researcher, hostmaster of AstroNet,
Sternberg Astronomical Institute, Moscow University (Russia)
Internet: oleg@sai.msu.su, http://www.sai.msu.su/~megera/
phone: +007(095)939-16-83, +007(095)939-23-83

Marc G. Fournier ICQ#7615664 IRC Nick: Scrappy
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org

Tom Lane

tgl@sss.pgh.pa.us

over 26 years ago

In reply to: The Hermit Hacker (#2)

Re: [HACKERS] How good is FreeBSD for postgres ?

The Hermit Hacker <scrappy@hub.org> writes:

All known problems...I've always run PostgreSQL under FreeBSD, and have
yet to notice a problem with a -release version of it...
The regression tests are "base Linux", and some of the generated error
messages are different between the systems and/or rounding is slighty
different...

The backend crashes that he's showing in some of the tests are not
known problems (to me anyway). Any ideas? Oleg, can you provide
debugger backtraces from the corefiles those crashes generate?

regards, tom lane

Import Notes

Reply to msg id not found: YourmessageofSun16May1999024648-0300Pine.BSF.4.05.9905160245250.47191-100000@thelab.hub.org | Resolved by subject fallback

Oleg Bartunov

oleg@sai.msu.su

over 26 years ago

In reply to: Tom Lane (#3)

Re: [HACKERS] How good is FreeBSD for postgres ?

I read across /usr/ports/databases/postgresql on my FreeBSD 3.1 release
elf machine and found a lot of patches to postgres 6.4.2
It seems they are didn't applied - I checked 6.5 cvs.
What is the status of this port ? Is there is a chance to apply them
into 6.5 release ?

Regards,

Oleg

On Sun, 16 May 1999, Tom Lane wrote:

Date: Sun, 16 May 1999 10:03:32 -0400
From: Tom Lane <tgl@sss.pgh.pa.us>
To: The Hermit Hacker <scrappy@hub.org>
Cc: Oleg Bartunov <oleg@sai.msu.su>, pgsql-hackers@postgreSQL.org
Subject: Re: [HACKERS] How good is FreeBSD for postgres ?

The Hermit Hacker <scrappy@hub.org> writes:

All known problems...I've always run PostgreSQL under FreeBSD, and have
yet to notice a problem with a -release version of it...
The regression tests are "base Linux", and some of the generated error
messages are different between the systems and/or rounding is slighty
different...

The backend crashes that he's showing in some of the tests are not
known problems (to me anyway). Any ideas? Oleg, can you provide
debugger backtraces from the corefiles those crashes generate?

regards, tom lane

Oleg Bartunov

oleg@sai.msu.su

over 26 years ago

In reply to: Oleg Bartunov (#4)

1 attachment(s)

Re: [HACKERS] How good is FreeBSD for postgres ?

On Sun, 16 May 1999, Tom Lane wrote:

Date: Sun, 16 May 1999 17:31:22 -0400
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Oleg Bartunov <oleg@sai.msu.su>
Subject: Re: [HACKERS] How good is FreeBSD for postgres ?

I'm new in freeBSD (FreeBSD nature.ru 3.1-RELEASE FreeBSD 3.1-RELEASE #2)
and it seems there is a problem with shared objects.

I'm not likely to be much help on that; it's probably a freeBSD-specific
issue. Try the mailing list; there are other people using freeBSD.
(Marc, I think, for one...)

Tom,

I probably understand what's going on. Most postgres developers still
use old a.out format while new versions switches to elf format
(what I'm actually using). I remember when it was happened with Linux
in 1995 and how many prolems were that time. I read freebsd mailing lists
and there are a lot of complaints about shared libs creating and usage.
Fortunately, I found that FreeBSD has nice porting policy and
found patches to Postgres 6.4.2. I was surprised they were never applied,
well, at least to current 6.5 cvs tree. Two patches I found very useful -
one to src/makefiles/Makefile.freebsd and another to
src/Makefile.shlib. I'm currently in process of rebuilding postgres
and will see. Hmm, I've lost connection to my server :-(
I could see in xterm that compilation and installation (with these patches)
passed OK and final message:

Thank you for choosing PostgreSQL, the most advanced open source database engine.

Anyway, FreeBSD guru, please examine these patches ! They are for
postgres 6.4.2 and FreeBSD 3.1 release.

Regards,

Oleg

regards, tom lane

Import Notes

Reply to msg id not found: 27503.926890282@sss.pgh.pa.us | Resolved by subject fallback

The Hermit Hacker

scrappy@hub.org

over 26 years ago

In reply to: Oleg Bartunov (#4)

Re: [HACKERS] How good is FreeBSD for postgres ?

Will look at them right now...

On Mon, 17 May 1999, Oleg Bartunov wrote:

I read across /usr/ports/databases/postgresql on my FreeBSD 3.1 release
elf machine and found a lot of patches to postgres 6.4.2
It seems they are didn't applied - I checked 6.5 cvs.
What is the status of this port ? Is there is a chance to apply them
into 6.5 release ?

Regards,

Oleg

On Sun, 16 May 1999, Tom Lane wrote:

Date: Sun, 16 May 1999 10:03:32 -0400
From: Tom Lane <tgl@sss.pgh.pa.us>
To: The Hermit Hacker <scrappy@hub.org>
Cc: Oleg Bartunov <oleg@sai.msu.su>, pgsql-hackers@postgreSQL.org
Subject: Re: [HACKERS] How good is FreeBSD for postgres ?

The Hermit Hacker <scrappy@hub.org> writes:

All known problems...I've always run PostgreSQL under FreeBSD, and have
yet to notice a problem with a -release version of it...
The regression tests are "base Linux", and some of the generated error
messages are different between the systems and/or rounding is slighty
different...

The backend crashes that he's showing in some of the tests are not
known problems (to me anyway). Any ideas? Oleg, can you provide
debugger backtraces from the corefiles those crashes generate?

regards, tom lane

_____________________________________________________________
Oleg Bartunov, sci.researcher, hostmaster of AstroNet,
Sternberg Astronomical Institute, Moscow University (Russia)
Internet: oleg@sai.msu.su, http://www.sai.msu.su/~megera/
phone: +007(095)939-16-83, +007(095)939-23-83

Marc G. Fournier ICQ#7615664 IRC Nick: Scrappy
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org

Oleg Bartunov

oleg@sai.msu.su

over 26 years ago

In reply to: The Hermit Hacker (#6)

Re: [HACKERS] How good is FreeBSD for postgres ?

On Mon, 17 May 1999, The Hermit Hacker wrote:

Date: Mon, 17 May 1999 00:35:00 -0300 (ADT)
From: The Hermit Hacker <scrappy@hub.org>
To: Oleg Bartunov <oleg@sai.msu.su>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, pgsql-hackers@postgreSQL.org
Subject: Re: [HACKERS] How good is FreeBSD for postgres ?

Will look at them right now...

I see you did a job ! Postgres 6.5 cvs works now under FreeBSD 3.1 release
(elf) as on my lovely Linux box ! Almost all regression tests passed
except float8, horology and geometry and rules (could somebody correct
pgsql <-> postgres in .out).

I think it would be useful to apply these fixes to 6.4.2 too.

Regards,
Oleg

PS.

10:05:47[nature]:/usr/home/postgres/cvs/pgsql/src/test/regress$ diff results/flo
at8.out expected/float8.out
190,191d189
< ERROR: floating point exception! The last floating point operation either exc
eeded legal ranges or was a divide by zero
< QUERY: SELECT '' AS bad, f.f1 ^ '1e200' from FLOAT8_TBL f;
192a191,192

QUERY: SELECT '' AS bad, f.f1 ^ '1e200' from FLOAT8_TBL f;
ERROR: pow() result is out of range

198,206c198
< bad| ?column?
< ---+--------------------
< | 1
< |7.39912306090513e-16
< | 0
< | 0
< | 1
< (5 rows)
<
---

ERROR: exp() result is out of range

223a216

ERROR: Input '10e-400' is out of range for float8

224a218

ERROR: Input '-10e-400' is out of range for float8

On Mon, 17 May 1999, Oleg Bartunov wrote:

I read across /usr/ports/databases/postgresql on my FreeBSD 3.1 release
elf machine and found a lot of patches to postgres 6.4.2
It seems they are didn't applied - I checked 6.5 cvs.
What is the status of this port ? Is there is a chance to apply them
into 6.5 release ?

Regards,

Oleg

On Sun, 16 May 1999, Tom Lane wrote:

Date: Sun, 16 May 1999 10:03:32 -0400
From: Tom Lane <tgl@sss.pgh.pa.us>
To: The Hermit Hacker <scrappy@hub.org>
Cc: Oleg Bartunov <oleg@sai.msu.su>, pgsql-hackers@postgreSQL.org
Subject: Re: [HACKERS] How good is FreeBSD for postgres ?

The Hermit Hacker <scrappy@hub.org> writes:

All known problems...I've always run PostgreSQL under FreeBSD, and have
yet to notice a problem with a -release version of it...
The regression tests are "base Linux", and some of the generated error
messages are different between the systems and/or rounding is slighty
different...

The backend crashes that he's showing in some of the tests are not
known problems (to me anyway). Any ideas? Oleg, can you provide
debugger backtraces from the corefiles those crashes generate?

regards, tom lane

_____________________________________________________________
Oleg Bartunov, sci.researcher, hostmaster of AstroNet,
Sternberg Astronomical Institute, Moscow University (Russia)
Internet: oleg@sai.msu.su, http://www.sai.msu.su/~megera/
phone: +007(095)939-16-83, +007(095)939-23-83

Marc G. Fournier ICQ#7615664 IRC Nick: Scrappy
Systems Administrator @ hub.org
primary: scrappy@hub.org secondary: scrappy@{freebsd|postgresql}.org

How good is FreeBSD for postgres ?

Attachments: