cloudlinux-collect: Collect System Load Statistics

Navigation:  Deprecated > OptimumCache >

cloudlinux-collect: Collect System Load Statistics

Previous pageReturn to chapter overviewNext page



cloudlinux-collectl Quick Start


Installing this package automatically starts system load statistics collection in background. cloudlinux-collectl package has no strict dependency on OptimumCache, thus the statistics is collected regardless of whether OptimumCache is installed or not. The aim of having this package pre-installed is to compare system performance before and after installing OptimumCache, thus to measure OptimumCache effectiveness.




# yum install cloudlinux-collect --enablerepo=cloudlinux-updates-testing


Note: cloudlinux-collectl will be installed automatically on optimumcache upgrade to 0.2-23.


Measure Web Site Response Time


cloudlinux-collectl can monitor response time for a configurable set of URLs.


Start monitoring new url:


# cloudlinux-collect --addurl <alias> <http://url>




# cloudlinux-collect --addurl localhost


Try ‘cloudlinux-collectl --help’ for more options.


To watch what is being collected


# cloudlinux-collect --test


Actual logs are compressed with gzip and kept in /var/log/optimumcache/collectl directory.


Statistics Being Collected in Details


To monitor what statistics are being collected, try command:


# cloudlinux-collect --test




Along with common statistics blocks as CPU, disk usage, inodes cached, there are two blocks of data to watch how effectively OptimumCache is functioning.


‘OPTIMUMCACHE DETAIL’ refers to data, which is similar to output of command


# optimumcache stat


csums:          4964 (99.9%)

             fetched        uncached            cached

inodes:          4967            31               4936      (99.4%)

size:          204177          131072           73104     (35.8%)

RAM:                8              4                4          (50.0%)


Particularly, the last column percent numbers shall match.


The next goes URLSTATTRACKER DETAIL block with url response time in milliseconds. Negative values here may pop up unexpectedly. Negative numbers are not milliseconds, but signal about http error response code for that specific url. For instance, -403 will signal for ‘Forbidden’ http error. As for -500 value, it signals not only for ‘Internal Server Error’, but can be displayed, when there is connection problem with the server, which is specified by the url.


Statistics Manual Configuration


URLSTATTRACKER DETAIL is the only statistics, which requires manual configuration. Upon clean installation, it has only url_localhost preconfigured:


# cloudlinux-collect --info


url shortname            url

--------------------         ---

localhost                 http://localhost/


To add another url for monitoring:


# cloudlinux-collect --addurl alt


To display urls being monitored list:


# cloudlinux-collect --info


url shortname          url

--------------------          ---


localhost                  http://localhost/


To skip url from being tracked run command:


# cloudlinux-collect --skip <url short name>


Running Statistics Daemon: collectl-cloudlinux


cloudlinux-collectl has got collectl package as a dependency. Initd script /etc/init.d/cloudlinux-collectl will automatically bring up another instance of collectl named ‘collectl-optimumcache’. collectl-optimumcache daemon instance has a separate config and does not interfere with other running pre-configure collectl daemon (if any).


As it was mentioned, collectl-optimumcache daemon starts automatically on package install, then on server restart events, kicked by regular Initd script /etc/init.d/cloudlinux-collectl. Thus, checking the daemon status, stop, restart is trivial:


# service cloudlinux-collect status

collectl-optimumcache (pid  1745) is running…


To start / stop :


# service cloudlinux-collect < start | stop >


Analyzing the Results


The statistics is being collected into files named %hostname%-%datetime%.raw.gz under directory /var/log/cloudlinux-collect


To convert those info format suitable for loading into Excel, LibreOffice Calc, another data mining tool, run the command:


# cloudlinux-collect --genplotfiles


         Generate fresh plot files in