On Friday 03 October 2003 19:52, Toad wrote: > > > Oh by the way, RoutingTime is a pretty bad estimator of CPU load. But > > > that problem is insignificant next to the others. > > > > I thought of a way to accept CPU load information: let > > the web server to accept the load as a parameter on a URL, > > and write a Perl script which finds the load and sends in > > the GET request. Or maybe a simpler protocol: listen on > > a port, accept a connection, and read load numbers from it. > > > > Then you don't have to do Java to C interfaces and use > > a different C implementation for each OS. > > Ugh. Python has a cross platform interface to obtain the current CPU > usage, across Windows, Linux, OS/X etc? Anyway we're not about to bundle > python AS WELL AS JAVA, unless there's a really good reason. > > Incidentally I have linux-specific code to parse /proc/stat and use that > in the load estimate, but including it would mean we'd have to include > the corresponding JNI windoze code.
Maybe I'm missing something, but why exactly is it useful to know the CPU load? What is the point? Freenet is so heavy anyway that it should probably run at nice levels between 15 and 19, certainly not lower. Otherwise even ridiculously powerful machines grind to a halt. Routing takes however long it takes. If the node starts getting bogged down with other unrelated load it will start getting routed around anyway. So what is the problem? Gordan _______________________________________________ Devl mailing list [EMAIL PROTECTED] http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl
