Обсуждение: Startup problem
Hello,
I have a problem running Postgres at startup. I use the start-scripts/linux script in /etc/rc.local. The script is executed on boot but the system ask a question :
" Your default context is user_u:system_r:unconfined_t Do you want to choose a different one? [n] "
If I do not answer the system wait several seconds and then continue boot (but I can't login except by connecting on ssh). Once I login I have to kill the start-scripts/linux process and launch it manually. When the script is manually launched the question is never asked
The answer to the question will always be no... I would like to know how to avoid this question, this machine will be rebooted remotely so I won't be able to type n !
The postgres version is 8.2.1 running on RedHat Enterprise Edition ES 4
I previously used a 7.6 version on Fedora Core 3 and there was no problem launching it on boot.
Thank you for your help!
Alex BECQUEREAU
--
SafeLogic
Hi, On Mon, 2007-01-22 at 12:20 +0100, Alexandre Becquereau wrote: > " Your default context is user_u:system_r:unconfined_t Do you want to > choose a different one? [n] " This is a SELinux issue. I haven't used SELinux until now; but you should either play with your SELinux policies or disable it. I'm sure some people may give a better answer. Regards, -- The PostgreSQL Company - Command Prompt, Inc. 1.503.667.4564 PostgreSQL Replication, Consulting, Custom Development, 24x7 support Managed Services, Shared and Dedicated Hosting Co-Authors: plPHP, ODBCng - http://www.commandprompt.com/
Вложения
"Alexandre Becquereau" <alex@safelogic.com> writes:
> I have a problem running Postgres at startup. I use the
> start-scripts/linux script in /etc/rc.local.
The one in our contrib/? That's pretty old, crufty, and unmaintained.
The one most people actually use on RPM-based systems is the one that
comes with our RPM distribution ... is there a reason you're not using
the RPM?
Anyway I would guess the problem comes from the fact that the contrib
script uses "su". The RPM script doesn't:
# For SELinux we need to use 'runuser' not 'su'
if [ -x /sbin/runuser ]
then
SU=runuser
else
SU=su
fi
and eventually
$SU -l postgres -c "$PGENGINE/postmaster -p '$PGPORT' -D '$PGDATA' ${PGOPTS} &" >> "$PGLOG" 2>&1 < /dev/null
but you really ought to adopt the whole script not just that one bit.
> I previously used a 7.6 version on Fedora Core 3 and there was no problem
> launching it on boot.
I believe FC3 didn't have SELinux security ... it certainly didn't have
it enabled by default.
regards, tom lane
Thank you, i replaced the su command by runuser and everything works
perfectly now!
Regards,
Alexandre BECQUEREAU
--
SafeLogic
www.safelogic.com
www.pgeep.com
-----Message d'origine-----
De : Tom Lane [mailto:tgl@sss.pgh.pa.us]
Envoyé : lundi 22 janvier 2007 17:10
À : Alexandre Becquereau
Cc : pgsql-admin@postgresql.org
Objet : Re: [ADMIN] Startup problem
"Alexandre Becquereau" <alex@safelogic.com> writes:
> I have a problem running Postgres at startup. I use the
> start-scripts/linux script in /etc/rc.local.
The one in our contrib/? That's pretty old, crufty, and unmaintained.
The one most people actually use on RPM-based systems is the one that comes
with our RPM distribution ... is there a reason you're not using the RPM?
Anyway I would guess the problem comes from the fact that the contrib script
uses "su". The RPM script doesn't:
# For SELinux we need to use 'runuser' not 'su'
if [ -x /sbin/runuser ]
then
SU=runuser
else
SU=su
fi
and eventually
$SU -l postgres -c "$PGENGINE/postmaster -p '$PGPORT' -D '$PGDATA'
${PGOPTS} &" >> "$PGLOG" 2>&1 < /dev/null
but you really ought to adopt the whole script not just that one bit.
> I previously used a 7.6 version on Fedora Core 3 and there was no
> problem launching it on boot.
I believe FC3 didn't have SELinux security ... it certainly didn't have it
enabled by default.
regards, tom lane
On Mon, Jan 22, 2007 at 11:10:27AM -0500, Tom Lane wrote: > > but you really ought to adopt the whole script not just that one bit. where is the is script to be found? Nothing extracts to /etc/init.d [stellr@stell postgresql]$ rpm -qlp postgresql-8.2.1-1PGDG.i686.rpm /usr/bin/clusterdb /usr/bin/createdb /usr/bin/createlang /usr/bin/createuser /usr/bin/dropdb /usr/bin/droplang /usr/bin/dropuser /usr/bin/pg_dump /usr/bin/pg_dumpall /usr/bin/pg_restore /usr/bin/psql /usr/bin/reindexdb /usr/bin/vacuumdb /usr/share/doc/postgresql-8.2.1 /usr/share/doc/postgresql-8.2.1/COPYRIGHT ...
found it: [stellr@stell postgresql]$ rpm -qlp postgresql-server-8.2.1-1PGDG.i686.rpm /etc/pam.d/postgresql /etc/rc.d/init.d/postgresql On Mon, Jan 22, 2007 at 01:20:21PM -0500, Ray Stell wrote: > On Mon, Jan 22, 2007 at 11:10:27AM -0500, Tom Lane wrote: > > > > but you really ought to adopt the whole script not just that one bit. > > > where is the is script to be found? Nothing extracts to /etc/init.d > > [stellr@stell postgresql]$ rpm -qlp postgresql-8.2.1-1PGDG.i686.rpm > /usr/bin/clusterdb > /usr/bin/createdb > /usr/bin/createlang > /usr/bin/createuser > /usr/bin/dropdb > /usr/bin/droplang > /usr/bin/dropuser > /usr/bin/pg_dump > /usr/bin/pg_dumpall > /usr/bin/pg_restore > /usr/bin/psql > /usr/bin/reindexdb > /usr/bin/vacuumdb > /usr/share/doc/postgresql-8.2.1 > /usr/share/doc/postgresql-8.2.1/COPYRIGHT > ... -- You have no chance to survive make your time.