You are not logged in.

#26 2011-07-30 14:22:55

Registered: 2011-04-12
Posts: 630

Re: Kernel IO schedulers: noop works better... why?

graysky wrote:

@unikum - yeah, but need to use iozone not hdparm...

EDIT:  I started a new thread so as not to contaminate this one with technical aspects of iozone.  Feel free to join the discussion, see results of the three schedulers, etc.

nice, i had thought about opening a new thread too, at one point but it obviously slipped my mind when posting.

i'm going to get iozone's extra stuff like the 3dplot running later today ~ i just have a few things to do first.



#27 2013-12-28 15:42:37

Registered: 2013-03-17
Posts: 75

Re: Kernel IO schedulers: noop works better... why?

Maybe it's possible to see a bigger difference using a mechanical disk. So what are the conclusions here?

That noop leaves the disk to schedule its IO. cfq and deadline offset IO scheduling to the CPU.

Last edited by archbawks (2013-12-28 15:48:50)


#28 2013-12-28 16:43:51

From: Pasadena, CA
Registered: 2009-07-13
Posts: 15,623

Re: Kernel IO schedulers: noop works better... why?

I do not think this thread is still viable.  It is more than two years old and much has changed in the kernel over that time.

Closing. … Bumping.22

Nothing is too wonderful to be true, if it be consistent with the laws of nature -- Michael Faraday
Sometimes it is the people no one can imagine anything of who do the things no one can imagine. -- Alan Turing
How to Ask Questions the Smart Way


Board footer

Powered by FluxBB