|
Plesk inside OpenVZ memory problems |
 |
 |
 |
|
Written by Tech Druid
|
|
Tuesday, 17 March 2009 09:58 |
|
I believe that the errors encountered below when installing Plesk inside an OpenVZ Virtual Environment (VE) were encountered as a result of memory allocation issues. Plesk recommends 256MB RAM minimum for 100 domains. However, I've found that this does not seem to be enough in a VE. On the hardware node (HN), here is the critical issue when installing inside a 256MB VE. # grep privvmpages /proc/user_beancounters resource held maxheld barrier limit failcnt
privvmpages 40347 74138 65536 73228 39 |
The fail count indicates a problem allocating memory during the plesk installation. Obviously I'm misunderstanding something about these settings. In any event, I'm starting over with 512MB RAM using the following settings.
vzctl set <veid> --save --privvmpages 131072:146456
vzctl set <veid> --save --vmguarpages 131072
vzctl set <veid> --save --oomguarpages 131072
vzctl set <veid> --save --shmpages 65536 |
Below are the vaious error messages encountered that brought me to implement the above settings.
Plesk Install Errors/Warnings.
WARNING!
During the start found some problems(see log file: /tmp/psa_9.0.1_fc8.build90090127.18_installing.090317.10.33.log)
Continue...
Trying to install chrooted environment... groups:
probably it will not work in chrooted acconts
WARNING!
During the register groups in chrooted environment found some problems(see log file: /tmp/psa_9.0.1_fc8.build90090127.18_installing.090317.10.33.log)
Continue...
done
===> Installing phpMyAdmin
Trying to create pmadb... done
done
===> Installing /usr/sbin/suexec replacement
Trying to backup original /usr/sbin/suexec... `/usr/sbin/suexec' -> `/usr/sbin/suexec.saved_by_psa'
done
Trying to reread IP addresses from network interfaces... done
done
WARNING!
During the default tempaltes cannot be created found some problems(see log file: /tmp/psa_9.0.1_fc8.build90090127.18_installing.090317.10.33.log)
Continue...
WARNING!
During the cannot update notifications. Please rerun /usr/local/psa/admin/plib/scripts/update_notifications.php manyally found some problems(see log file: /tmp/psa_9.0.1_fc8.build90090127.18_installing.090317.10.33.log)
Continue...
Trying to tune Admin apache... cp: cannot stat `/usr/local/psa/admin/conf/httpsd.conf.def': No such file or directory
cat: /usr/local/psa/admin/conf/httpsd.conf.def-pre_daemon_tune: No such file or directory
cp: cannot stat `/usr/local/psa/admin/conf/httpsd.conf': No such file or directory
cat: /usr/local/psa/admin/conf/httpsd.conf-pre_daemon_tune: No such file or directory
Trying to executing mail_restore to synchronize mail server settings and Plesk Database... done
Shutting down postfix: [ OK ]
Starting postfix: [ OK ]
Trying to start Plesk...
Stopping Plesk... done
Starting xinetd service... done
Starting named service... failed
Starting mysqld service... failed
Starting postgresql service... not installed
Starting psa-spamassassin service... not installed
Plesk: Starting Mail Server... already started
Starting Plesk... done
Starting drwebd service... not installed
Trying to start service httpd... httpd (pid 1849 1839) is running...
done
<snip>
Installing the package psa-awstats-configurator-1.0.0-fc8.build90090127.18.noarch (b 0x9d6db48 source=0x9499c00 P:1 R:3)
ERROR: Failed to execute a package installation command.
Failed to execute an external command: fork operation failed.
Warning! Not all packages were installed.
Please check if your system services are operable.
Please, resolve the above problem and try installing the packages again.
If you cannot resolve the problem on your own, contact product technical support for assistance.
/var/log/messages
Mar 17 10:33:28 peck named[32359]: starting BIND 9.5.0-P2 -u named -c /etc/named.conf -u named -t /var/named/run-root
Mar 17 10:33:28 peck named[32359]: found 2 CPUs, using 2 worker threads
Mar 17 10:33:28 peck named[32359]: ./main.c:488: unexpected error:
Mar 17 10:33:28 peck named[32359]: isc_taskmgr_create() failed: no available threads
Mar 17 10:33:28 peck named[32359]: create_managers() failed: unexpected error
Mar 17 10:33:28 peck named[32359]: exiting (due to early fatal error)
Mar 17 10:33:28 peck named[32421]: starting BIND 9.5.0-P2 -u named -c /etc/named.conf -u named -t /var/named/run-root
Mar 17 10:33:28 peck named[32421]: found 2 CPUs, using 2 worker threads
Mar 17 10:33:28 peck named[32421]: ./main.c:488: unexpected error:
Mar 17 10:33:28 peck named[32421]: isc_taskmgr_create() failed: no available threads
Mar 17 10:33:28 peck named[32421]: create_managers() failed: unexpected error
Mar 17 10:33:28 peck named[32421]: exiting (due to early fatal error)
Mar 17 10:33:41 peck named[1799]: starting BIND 9.5.0-P2 -u named -c /etc/named.conf -u named -t /var/named/run-root
Mar 17 10:33:41 peck named[1799]: found 2 CPUs, using 2 worker threads
Mar 17 10:33:41 peck named[1799]: ./main.c:488: unexpected error:
Mar 17 10:33:41 peck named[1799]: isc_taskmgr_create() failed: no available threads
Mar 17 10:33:41 peck named[1799]: create_managers() failed: unexpected error
Mar 17 10:33:41 peck named[1799]: exiting (due to early fatal error)
|
