11-21-2009, 07:16 PM
the ultimate question is, if your server runs without any drop in a long time measurement over a month :-)
but this (more or less) not about the time interval of the measurement (5 minutes, 1 hour, 1 month, ...) but about the number of measurements. the fast measurement uses 300 measurements, the normal 720 (and thus is more precise!) and the long time measurement many more...
beware that 0 drops in a measurement don't mean your server doesn't drop. you can only say that it will drop in less then 1 of 300 (or 720 or what ever is your number of measurements) frames. so to know how your server runs really you need to make as much as possible measurements. easiest way is a long-time measurement, running a measurement longer than a week will also eliminate all daytime and day-of-week effects (from the server load)...
but this (more or less) not about the time interval of the measurement (5 minutes, 1 hour, 1 month, ...) but about the number of measurements. the fast measurement uses 300 measurements, the normal 720 (and thus is more precise!) and the long time measurement many more...
beware that 0 drops in a measurement don't mean your server doesn't drop. you can only say that it will drop in less then 1 of 300 (or 720 or what ever is your number of measurements) frames. so to know how your server runs really you need to make as much as possible measurements. easiest way is a long-time measurement, running a measurement longer than a week will also eliminate all daytime and day-of-week effects (from the server load)...
http://www.fpsmeter.org
http://wiki.fragaholics.de/index.php/EN:Linux_Optimization_Guide (Linux Kernel HOWTO!)
Do not ask technical questions via PM!
http://wiki.fragaholics.de/index.php/EN:Linux_Optimization_Guide (Linux Kernel HOWTO!)
Do not ask technical questions via PM!