Danke Ninos, ich hab der DB 16 GB gegeben von 32, der RAM Verbrauch ist relativ gering 50% ...
swapp = 0
diese Meldung taucht manchmal beim arbeiten im terminal auf, scheint aber damit nicht zusammenzuhängen:
Posts by fulltilt
-
-
habe gestern einen Server Umzug durchgeführt von einem 16 GB auf einen 8 core 32 GB dedi
die Load auf dem neuen System ist deutlich höher, habe mal die my.cnf auf innodb umgessstellt und auch ordentlich RAM zur Verfügung gestellt - nix ...
auf dem Server laufen nur 3 Webs ... die php-fcgi Prozesse sind zwischen 60 und 90% (memcached und APC cache sind auch installiert)
Irgendwie habe ich den Eindruck das Squeeze besser arbeitet im Vergleich zu Wheezy, wollte in Kürze auch andere Server upgraden, aber das lasse ich erst einmal auf Squeeze laufen denn diese haben ca. 250 Webs und dieser hier kommt mit dreien schon an seine Grenzenkönnte das an den Platten liegen?
sieht aber normal aus ... oderCode- # hdparm -tT --direct /dev/sda
- /dev/sda:
- Timing O_DIRECT cached reads: 490 MB in 2.00 seconds = 244.61 MB/sec
- Timing O_DIRECT disk reads: 504 MB in 3.20 seconds = 157.33 MB/sec
- # hdparm -tT --direct /dev/sdb
- /dev/sdb:
- Timing O_DIRECT cached reads: 484 MB in 2.01 seconds = 241.24 MB/sec
- Timing O_DIRECT disk reads: 506 MB in 3.01 seconds = 168.14 MB/sec
-
you're right, as mostly always
I guess I forgot to migrate the fcgi folder and after the first stop the status from imscp tables have been set to whatever ..
thanks Nuxwin -
thank you, but this is not ispcp to imscp ...
I use IMSCP 1.1.5 on both servers, the only difference is the debian version (old server squeeze - new one wheezy).
But I will re-check the database status for all tables and retry -
thanks Nuxwin, sounds good
but what should I do? -
same again, seems it is not possible to migrate from one to another server with 1.1.5.
strange ... seems the imscp daemon is been killed during the process
but a reboot did not helped, also removed the first user and group which has been created
connection closed by remote host ... during installs for apache_fcgid_tasks
also my (debian kde) konsole window have been killed after the error message, so I have to close and re-run konsole -
after deleting the fcgi folder it runs until the customer rebuild:
Code- [Tue Apr 1 18:40:10 2014] [debug] iMSCP::Ext2Attributes::__ANON__: Clearing 'Immutable' extended attribute on /var/www/virtual/domain.com
- [Tue Apr 1 18:40:10 2014] [debug] iMSCP::Execute::execute: Executing command: /usr/bin/pkill -KILL -u vu2006 ; /usr/sbin/usermod -c 'i-MSCP Web User' -d '/var/www/virtual/domain.com' -l vu2004 -m -s '/bin/false' vu2004
- [Tue Apr 1 18:40:10 2014] [debug] iMSCP::Execute::getExitCode: External command exited with value 6
- [Tue Apr 1 18:40:10 2014] [debug] Modules::User::add: usermod: user 'vu2004' does not exist
-
if I start perl imscp-autoinstall -d it stops during step 4 server installations with:
connection to xxx.xxx.xxx.xx closed by remote hostthe database have been imported from Squeeze to Wheezy
the only config files I use:
/etc/imscp/imscp.conf
/etc/imscp/imscp-db-leys
and debian.cnf -
which option should be used for the migration process (same imscp version to a new server) ...
after syncing the files and database, build or install ?Regards
-