Обсуждение: error in postgresql

pls check for the error and tell me if any solution can be come upon.
TIA
Sharmad
Script started on Tue Oct 30 10:24:36 2001
postgres@linux:~ > pg_dumpall > alllms_30-10-2001.bak
SQL query to dump the contents of Table 'common_to_three' did not execute.  Explanation from backend: 'pqReadData() --
backendclosed the channel unexpectedly. 
    This probably means the backend terminated abnormally
    before or while processing the request.
'.
The query was: 'COPY "common_to_three" TO stdout;
'.
pg_dump failed on lms, exiting
postgres@linux:~ >
Script done on Tue Oct 30 10:24:48 2001

sharmad@goatelecom.com writes:
> postgres@linux:~ > pg_dumpall > alllms_30-10-2001.bak
> SQL query to dump the contents of Table 'common_to_three' did not execute.  Explanation from backend: 'pqReadData()
--backend closed the channel unexpectedly.
 
>     This probably means the backend terminated abnormally
>     before or while processing the request.
> '.
> The query was: 'COPY "common_to_three" TO stdout;

Hm.  The only cases I can recall of backend crashes during COPY OUT
were caused by corrupted data.  Do you get similar failures if you
do "SELECT * FROM common_to_three"?

You could investigate more closely by using gdb to look at the core
file left by the crashed backend.  A backtrace would give us at least
some idea of the nature of the problem.
        regards, tom lane

Sharmad Naik <sharmad@goatelecom.com> writes:
>     When I gave a search for core file I didn't seems to
> exist.pg_dump didnot create the core file

pg_dump isn't coring, it's the backend.  Backend core files normally
appear in $PGDATA/base/whicheverdb/.  If you don't see one there,
it's likely because you started the postmaster under "ulimit -c 0"
(this seems to be the default for init-script-started processes on
at least some Linux distros :-().  Restart it with "ulimit -c unlimited"
and try again.
        regards, tom lane

Hi,Firstly I'm Sorry for the delay in replying.When I gave a search for core file I didn't seems to
exist.pg_dump didnot create the core filewell if i tried to even query the table of the database it 
would directly close  the connection with the backend giving the
prompt like #> and displaying a similar message shown below.Pls direct me.

-TIA
Sharmad
Pls cc a copy to sharmad@softhome.net

On Tue, Oct 30, 2001 at 04:46:56PM -0500, Tom Lane wrote:
> sharmad@goatelecom.com writes:
> > postgres@linux:~ > pg_dumpall > alllms_30-10-2001.bak
> > SQL query to dump the contents of Table 'common_to_three' did not execute.  Explanation from backend: 'pqReadData()
--backend closed the channel unexpectedly.
 
> >     This probably means the backend terminated abnormally
> >     before or while processing the request.
> > '.
> > The query was: 'COPY "common_to_three" TO stdout;
> 
> Hm.  The only cases I can recall of backend crashes during COPY OUT
> were caused by corrupted data.  Do you get similar failures if you
> do "SELECT * FROM common_to_three"?
> 
> You could investigate more closely by using gdb to look at the core
> file left by the crashed backend.  A backtrace would give us at least
> some idea of the nature of the problem.


-- 
Feel free to contact me (flames about my english and the useless of this
driver will be redirected to /dev/null, oh no, it's full...).    (Michael Beck, describing the PC-speaker sound
device)
---------------------
__    _  _      _  _
|_|_||_||_||\/||_|| \
_|| || || \|  || ||_/"What we cannot speak about we must pass over in silence."       Ludwig Wittgenstein (1889-1951),
Austrianphilosopher
 

Hi Tom,It's not there. Now do i say postmaster -D <directory name of dbase> ulimit -c 0
Is this right?
-TIA
Sharmad

On Sun, Nov 11, 2001 at 12:00:39PM -0500, Tom Lane wrote:
> Sharmad Naik <sharmad@goatelecom.com> writes:
> >     When I gave a search for core file I didn't seems to
> > exist.pg_dump didnot create the core file
> 
> pg_dump isn't coring, it's the backend.  Backend core files normally
> appear in $PGDATA/base/whicheverdb/.  If you don't see one there,
> it's likely because you started the postmaster under "ulimit -c 0"
> (this seems to be the default for init-script-started processes on
> at least some Linux distros :-().  Restart it with "ulimit -c unlimited"
> and try again.
> 
>             regards, tom lane

-- 
Feel free to contact me (flames about my english and the useless of this
driver will be redirected to /dev/null, oh no, it's full...).    (Michael Beck, describing the PC-speaker sound
device)
---------------------
__    _  _      _  _
|_|_||_||_||\/||_|| \
_|| || || \|  || ||_/"What we cannot speak about we must pass over in silence."       Ludwig Wittgenstein (1889-1951),
Austrianphilosopher