[ Comments below, in line ] On Wednesday 06 January 2010 at 11:34 am, Tobias Oetiker penned about "Re: [vbox-users] vbox performance issues verified with benchmarking"
> Hi Pablo, Hi Tobi, > the disk is a raid6 system with ext3 ... its performance is not the > issue for our application ... 'k > what is killing me is the slowdown effect ... it causes windows to > grind to a halt, requiering a reboot .. very unplesant ... Any `strace' information on `before' and `after'? This may prove to be beneficial ... or not! :) > ps. apropos fs performance, check out > http://insights.oetiker.ch/linux/fsopbench/ re: above; You're right, a true indicator are multiple readers and writers simulating _your_ environment. Sizing IO needs to consider in-rush (r) and out-rush (w) to the subsystem, otherwise you end up with outliers as noted. And Linux gives you many knobs to tune which also adds to the complexity ... as you noted above. :) Cheers, -- Pablo Sanchez - Blueoak Database Engineering, Inc Ph: 819.459.1926 Fax: 760.860.5225 (US) _______________________________________________ vbox-users mailing list vbox-users@virtualbox.org http://vbox.innotek.de/mailman/listinfo/vbox-users
