Comments on: I/O scheduler and SSD: part 2! http://www.alphatek.info/2009/02/02/io-scheduler-and-ssd-part-2/ Random rants and tips about Fedora, CentOS, Maemo and other things I care about Mon, 19 Jul 2010 22:10:25 +0000 hourly 1 http://wordpress.org/?v=3.0 By: jimn http://www.alphatek.info/2009/02/02/io-scheduler-and-ssd-part-2/comment-page-1/#comment-7697 jimn Tue, 15 Sep 2009 16:52:12 +0000 http://www.alphatek.info/?p=307#comment-7697 for those of us with spindles still, NCQ disk access coupled with elevator=noop is suggested to allow the hardware to do fine grained elevator control which is defeated by IO scheduling at the kernel layer. fwiw. i'd stick with noop on gp, for SSD too, however, for anything but server loads. for those of us with spindles still, NCQ disk access coupled with elevator=noop is suggested to allow the hardware to do fine grained elevator control which is defeated by IO scheduling at the kernel layer. fwiw. i’d stick with noop on gp, for SSD too, however, for anything but server loads.

]]> By: mAAddog http://www.alphatek.info/2009/02/02/io-scheduler-and-ssd-part-2/comment-page-1/#comment-3698 mAAddog Sat, 07 Feb 2009 04:34:25 +0000 http://www.alphatek.info/?p=307#comment-3698 Well it will make a good part 3 if you ever do :) Well it will make a good part 3 if you ever do :)

]]> By: Steven http://www.alphatek.info/2009/02/02/io-scheduler-and-ssd-part-2/comment-page-1/#comment-3682 Steven Thu, 05 Feb 2009 07:26:19 +0000 http://www.alphatek.info/?p=307#comment-3682 Honestly, I don't have much time for that right now. Honestly, I don’t have much time for that right now.

]]> By: mAAddog http://www.alphatek.info/2009/02/02/io-scheduler-and-ssd-part-2/comment-page-1/#comment-3678 mAAddog Wed, 04 Feb 2009 22:03:05 +0000 http://www.alphatek.info/?p=307#comment-3678 Nice article, though NOOP is not much of a scheduler. Which is i guess, why you were testing it :D Since you posted a followup, what about another one with "elevator=deadline" as well as "echo 1 > /sys/block/sdX/queue/iosched/fifo_batch" to reduce read/write grouping. Cheers mAAddog Nice article, though NOOP is not much of a scheduler. Which is i guess, why you were testing it :D

Since you posted a followup, what about another one with “elevator=deadline” as well as “echo 1 > /sys/block/sdX/queue/iosched/fifo_batch” to reduce read/write grouping.

Cheers
mAAddog

]]> By: Nicolas http://www.alphatek.info/2009/02/02/io-scheduler-and-ssd-part-2/comment-page-1/#comment-3657 Nicolas Tue, 03 Feb 2009 14:06:56 +0000 http://www.alphatek.info/?p=307#comment-3657 To be not cpu limited, try to bench only the decompression speed of the big tar.gz file. Usualy random write are quite bad on cheap SSD (!= intel) that should be a good test. (random write) To be not cpu limited, try to bench only the decompression speed of the big tar.gz file. Usualy random write are quite bad on cheap SSD (!= intel) that should be a good test. (random write)

]]>