On Mon, Aug 13, 2007 at 09:29:04PM +0200, Artur Grabowski wrote: > Joachim Schipper <[EMAIL PROTECTED]> writes: > > > Well, aside from the 'the load average is useless anyway' comments, a > > load average of 1 is not necessarily unacceptable on a multi-processor > > machine. (It means one CPU is constantly working.) > > No, no, no and NO. > > Please. If you don't understand load averages, do not try to interpret them.
Which is the whole problem with load averages, in a nutshell. > It means that whenever the once a second event in the kernel happens, > something was either in the run queue or the system made a guess that > something might end up in the run queue within a second. That guess is > not always correct. If you have something that wakes up once a second > without doing anything, it will add to the load average. Seeing an unusual load average means something unusual is happening. This may not be a bad thing at all. In fact, it almost never is a bad thing when the normal is 0.20 and you see 1.70. The *stat tools and top quickly show what's happening, and it usually just means something's actually doing something on an otherwise quiet box. Good! But if anyone really wants to monitor their systems, I'd advise them to use different and more indicative metrics. Load average *does* have a meaning, but it doesn't tell you much about what's going on. At best it's just a somewhat capricious red flag to look elsewhere. Why not start by looking elsewhere instead? Monitoring tools are available in ports/packages and some of them even make nice graphs, which are dead useful in spotting trends and abnormal behavior. -- Darrin Chandler | Phoenix BSD User Group | MetaBUG [EMAIL PROTECTED] | http://phxbug.org/ | http://metabug.org/ http://www.stilyagin.com/ | Daemons in the Desert | Global BUG Federation
