Messages in this thread |  | | | Date | Fri, 25 Sep 2009 16:46:34 +0200 (CEST) | | From | Tobias Oetiker <> | | Subject | Re: [Fwd: [RFC] cfq: adapt slice to number of processes doing I/O (v2.1)] |
| |
Hi Corrado,
> > Have you tested buffered write, multi-threads? > Yes. One interesting test was done by Tobias Oetiker > (http://oss.oetiker.ch/optools/wiki/fsopbench) on his Areca HW Raid6 > with ext3. He put target_latency to 200ms, and used a lower > slice_idle, to match his faster disks.
I have since updated my benchmark program:
http://oss.oetiker.ch/optools/wiki/FsOpBench
it provides an even better competition model. It now works on a 20GB artificial file tree with a home directory like file distribution. Results are quite striking ... (not necessarily encouraging).
results are quite amazing ... I am still collecting ... but maybe some of you want to try this too ...
cheers tobi
-- Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland http://it.oetiker.ch tobi@oetiker.ch ++41 62 775 9902 / sb: -9900
|  |