That CPU usage will be wildly inaccurate for bursty tasks though I can see the benefit to logging memory consumption once every minute.
Ideally, you'd want to include something which somehow asks the kernel for auditing data and records each process's cumulative CPU time and wall time on exit (as well as any memory statistics offered) so you don't miss any short-lived processes. On 14-05-24 03:05 PM, Mark Constable wrote: > On 25/05/14 03:58, Stephan Sokolow wrote: >> I can't get a meaningful output via your method since many of them >> aren't kept resident and your method doesn't distinguish between >> long-running processes and resource-heavy ones. > > But it does provide a snapshot of what is currently in use sorted by > cpu usage vs what one may or may not want to use. It's also in a format > that could be parsed and if there was a snapshot every 1 or 5 minutes > into a SQLite database then it would be more useful averaged over a > week or month(s). Times that by dozens of regular contributors then in > a few months we would have an accurate idea of what folks actually use. > > > ------------------------------------------------------------------------------ > "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE > Instantly run your Selenium tests across 300+ browser/OS combos. > Get unparalleled scalability from the best Selenium testing platform available > Simple to use. Nothing to install. Get started now for free." > http://p.sf.net/sfu/SauceLabs > ------------------------------------------------------------------------------ "Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE Instantly run your Selenium tests across 300+ browser/OS combos. Get unparalleled scalability from the best Selenium testing platform available Simple to use. Nothing to install. Get started now for free." http://p.sf.net/sfu/SauceLabs _______________________________________________ Lxde-list mailing list Lxde-list@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lxde-list
