Обсуждение: troubles with postgresql 7.0.2
After working a while with postgresql 7.0.2 from redhat 7.0, with upgrates
at glibc 2.2.2 and a few other improvements, i have kernel 2.2.17 from rh
7.0 updates.
And today my postmaster died with :
The Postmaster has informed me that some other backend died abnormally and
possibly corrupted shared memory.
I have rolled back the current transaction and am going to
terminate your database system connection and exit.
and i wasn't able to start again.
/usr/bin/postmaster: reaping dead processes...
/usr/bin/postmaster: CleanupProc: pid 1483 exited with status 139
Server process (pid 1483) exited with status 139 at Thu May 3 10:29:59
2001
Terminating any active server processes...
Server processes were terminated at Thu May 3 10:29:59 2001
Reinitializing shared memory and semaphores
010503.10:29:59.379 [1477] shmem_exit(0)
binding ShmemCreate(key=52e389, size=1104896)
010503.10:29:59.595 [1484] DEBUG: Data Base System is starting up at Thu
May
3 10:29:59 2001
010503.10:29:59.596 [1484] DEBUG: Data Base System was interrupted being
in production at Thu May 3 10:25:58 2001
010503.10:29:59.599 [1484] DEBUG: Data Base System is in production
state at Thu May 3 10:29:59 2001
010503.10:29:59.599 [1484] proc_exit(0)
010503.10:29:59.600 [1484] shmem_exit(0)
010503.10:29:59.600 [1484] exit(0)
/usr/bin/postmaster: reaping dead processes...
this is the error message now, and i have no idee how can i resolv this.
i use /usr/bin/pg_ctl -o "-i -d 5" -D /[datadir] start &
Diana Cionoiu
> After working a while with postgresql 7.0.2 from redhat 7.0, with upgrates
> at glibc 2.2.2 and a few other improvements, i have kernel 2.2.17 from rh
> 7.0 updates.
> And today my postmaster died with :
>
> The Postmaster has informed me that some other backend died abnormally and
> possibly corrupted shared memory.
> I have rolled back the current transaction and am going to
> terminate your database system connection and exit.
>
> and i wasn't able to start again.
>
> /usr/bin/postmaster: reaping dead processes...
> /usr/bin/postmaster: CleanupProc: pid 1483 exited with status 139
> Server process (pid 1483) exited with status 139 at Thu May 3 10:29:59
> 2001
> Terminating any active server processes...
> Server processes were terminated at Thu May 3 10:29:59 2001
> Reinitializing shared memory and semaphores
> 010503.10:29:59.379 [1477] shmem_exit(0)
> binding ShmemCreate(key=52e389, size=1104896)
> 010503.10:29:59.595 [1484] DEBUG: Data Base System is starting up at Thu
> May
> 3 10:29:59 2001
> 010503.10:29:59.596 [1484] DEBUG: Data Base System was interrupted being
> in production at Thu May 3 10:25:58 2001
> 010503.10:29:59.599 [1484] DEBUG: Data Base System is in production
> state at Thu May 3 10:29:59 2001
> 010503.10:29:59.599 [1484] proc_exit(0)
> 010503.10:29:59.600 [1484] shmem_exit(0)
> 010503.10:29:59.600 [1484] exit(0)
> /usr/bin/postmaster: reaping dead processes...
>
> this is the error message now, and i have no idee how can i resolv this.
> i use /usr/bin/pg_ctl -o "-i -d 5" -D /[datadir] start &
>
> Diana Cionoiu
Escuze me, I have find out more, the database still working, but take some
time after that query. i have 2 querys from php.
$mumu = pg_exec ($conn, "SELECT idl,datapromulgari,tipdocument,denumirea
from ldi where categoria=" . $categorie. " and denumirea LIKE '%" .
$denumirea . "%' and substr(dataemiterii,1,2) LIKE '%" . $luna ."%' and
substr(dataemiterii,3,2) LIKE '%" . $zi . "%' and substr(dataemiterii,5,4)
LIKE '%" . $an . "%' and descriere LIKE '%" . $descriere . "%' and
organemitent LIKE '%" . chop($organ) . "%' and publicat LIKE '%" .
chop($tippub) ."%' and tipdocument LIKE '%" . chop($tipdoc). "%' order by
to_date(dataemiterii,'MMDDYYYY') ASC LIMIT " . $nrlegi . "," . $nr );
first is working.
$mimuv = pg_exec ($conn, "SELECT idl from ldi where categoria=" .
$categorie. " and denumirea LIKE '%" . $denumirea . "%' and
substr(dataemiterii,1,2) LIKE '%" . $luna ."%' and
substr(dataemiterii,3,2) LIKE '%" . $zi . "%' and substr(dataemiterii,5,4)
LIKE '%" . $an . "%' and descriere LIKE '%" . $descriere . "%' and
organemitent LIKE '%" . chop($organ) . "%' and publicat LIKE '%" .
chop($tippub) ."%' and tipdocument LIKE '%" . chop($tipdoc). "%'" );
second executed after first, give me the error message.
Diana
Diana Cionoiu <diana@softwaresolutions.ro> writes: > $mimuv = pg_exec ($conn, "SELECT idl from ldi where categoria=" . > $categorie. " and denumirea LIKE '%" . $denumirea . "%' and > substr(dataemiterii,1,2) LIKE '%" . $luna ."%' and > substr(dataemiterii,3,2) LIKE '%" . $zi . "%' and substr(dataemiterii,5,4) > LIKE '%" . $an . "%' and descriere LIKE '%" . $descriere . "%' and > organemitent LIKE '%" . chop($organ) . "%' and publicat LIKE '%" . > chop($tippub) ."%' and tipdocument LIKE '%" . chop($tipdoc). "%'" ); > second executed after first, give me the error message. Possibly you are running into the LIKE-at-end-of-memory bug identified last July (see rev 1.37 at http://www.postgresql.org/cgi/cvsweb.cgi/pgsql/src/backend/utils/adt/like.c). Please update to 7.0.3 ... or even better, 7.1, but 7.0.3 wouldn't force you into a database dump/reload ... and see if you still see the problem. regards, tom lane
Diana Cionoiu <diana@softwaresolutions.ro> writes:
> And today my postmaster died with :
> The Postmaster has informed me that some other backend died abnormally and
> possibly corrupted shared memory.
> I have rolled back the current transaction and am going to
> terminate your database system connection and exit.
Backtrace from crashed backend's core file, please?
> and i wasn't able to start again.
Why not? The log extract you give shows the postmaster recovering
perfectly normally.
regards, tom lane
Diana Cionoiu <diana@softwaresolutions.ro> writes: > > After working a while with postgresql 7.0.2 from redhat 7.0, with upgrates > > at glibc 2.2.2 and a few other improvements, i have kernel 2.2.17 from rh > > 7.0 updates. > > And today my postmaster died with : Try the postgresql packages from Red Hat Linux 7.1. -- Trond Eivind Glomsrød Red Hat, Inc.
> Diana Cionoiu <diana@softwaresolutions.ro> writes: > > $mimuv = pg_exec ($conn, "SELECT idl from ldi where categoria=" . > > $categorie. " and denumirea LIKE '%" . $denumirea . "%' and > > substr(dataemiterii,1,2) LIKE '%" . $luna ."%' and > > substr(dataemiterii,3,2) LIKE '%" . $zi . "%' and substr(dataemiterii,5,4) > > LIKE '%" . $an . "%' and descriere LIKE '%" . $descriere . "%' and > > organemitent LIKE '%" . chop($organ) . "%' and publicat LIKE '%" . > > chop($tippub) ."%' and tipdocument LIKE '%" . chop($tipdoc). "%'" ); > > > second executed after first, give me the error message. > > Possibly you are running into the LIKE-at-end-of-memory bug identified > last July (see rev 1.37 at > http://www.postgresql.org/cgi/cvsweb.cgi/pgsql/src/backend/utils/adt/like.c). > Please update to 7.0.3 ... or even better, 7.1, but 7.0.3 wouldn't force > you into a database dump/reload ... and see if you still see the > problem. Thank you very much. Was the time to use a newer postgresql, i will start to port everything now on 7.1, witch i have it in tests since 2 weeks ago. > regards, tom lane Diana Cionoiu
> Possibly you are running into the LIKE-at-end-of-memory bug identified > last July (see rev 1.37 at > http://www.postgresql.org/cgi/cvsweb.cgi/pgsql/src/backend/utils/adt/like.c). > Please update to 7.0.3 ... or even better, 7.1, but 7.0.3 wouldn't force > you into a database dump/reload ... and see if you still see the > problem. Thx alot Tom, i have upgrade today at postgresql 7.1, and after a few problems with ssl, witch is not very well documented, it rocks the old one, was resolv 3 of my problems, speed, Like problem, and 8k field problem. > regards, tom lane Diana Cionoiu