- Plesk for Linux
Health Monitor shows high CPU usage after reboot:
CONFIG_TEXT: Total usage 8998100.9% used (?)
The following warning may appear when upgrading from older Plesk versions to Plesk 11.x or 12.x:
PLESK_ERROR: Parallels Panel pre-upgrade check... WARNING: If you leave the default value of the "Interval" parameter in the /etc/sw-collectd/collectd.conf, sw-collectd may heavily load the system.
sw-collectd process could consume a lot of CPU in older Plesk versions as it used /proc filesystem to get information about processes.
Increase "interval" value in Health Monitor config:
- Connect to the server via SSH
Note: If direct SSH/RDP access to the server is not possible, contact server administrator for further assistance.
Stop sw-collectd service:
# /etc/init.d/sw-collectd stop
/etc/sw-collectd/collectd.conffile to include/change 'Interval 300' line at top level. Replace '300' with desired interval in seconds (recommended '600'):
CONFIG_TEXT: BaseDir "/usr/local/psa/var/health/data"
Hostnamedirectives in the configuration file. From them a path to the statistics storage is assembled.
After changing interval, delete the collected statistics:
# rm -rf /usr/local/psa/var/health/data/localhost/*
Start sw-collectd service:
# /etc/init.d/sw-collectd restart